Explanations - cdn-links.lww.com€¦ · Web viewNEW ZEALAND - TABLETS ONLY. OMAN. PAKISTAN. PANANA. QATAR. SALVADOR. SINGAPOUR - TABLETS ONLY. SWITZERLAND. THE BARBADOS - TABLETS

Supplemental Table 3. Pain Group Evidence Summaries and Evidence to Decision Tables

Question: Adjunctive nefopam compared to no adjunctive nefopam for ICU pain management (5.2) Quality assessment № of patients Effect

Quality Importance№ of studie

s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other considerations

adjunctive nefopam

no adjunctive nefopam

Relative

(95% CI)

Absolute

(95% CI)

VAS Score with movement

1 randomised trials

serious a not serious serious b serious c none Overall no difference in Pain assessment with or without nefopam: at rest and during movement at 12, 24, 36, 48 and 72 h after surgery using a visual analogue scale (VAS) Pain with movement at 48h was the only time point for which the difference was statistically significant. All others were non-significant.

⨁◯◯◯VERY LOW

CRITICAL

CI: Confidence interval; MD: Mean differenceExplanationsa. Randomization method not described, no intention to treat with 24/300 loss to follow-up, blinding described b. Only included CVICU patients in this study, however analgesic efficacy may be broadly applicable to all ICU patients c. Confidence Intervals do not cross z ero (no effect) - however small sample siz e hard to draw conclusion on nefopam effects QuestionShould adjunctive nefopam vs. no adjunctive nefopam be used for ICU pain management (5.2)?

POPULATION: ICU pain management (5.2) BACKGROUND:

INTERVENTION:

adjunctive nefopam

COMPARISON: no adjunctive nefopam

MAIN OUTCOMES:

VAS Score with movement;

SETTING:

PERSPECTIVE:

JUDGEMENT RESEARCH EVIDENCE ADDITIONAL CONSIDERATIONS

PRO

BLE

M

Is the problem a priority?○ No○ Probably no○ Probably yes● Yes

○ Varies○ Don't know

It is top priority to investigate the effectiveness/side effects of Adjunctive Non opioid analgesics to opioids compared to opioids alone in order to decrease opioids related side effects that are especially at risk in ICU patients:- respiratory depression- neurological impairment (decreased vigilance, delirium)- ileus, nausea vomiting, etc...Kim et al. 2014 [1]n=276 patients scheduled to undergo cardiac surgery were randomly assigned between threePCA groups (92 patients per group): nefopam, fentanyl or nefopam + fentanyl.Each drug was given continuously + in PCA mode.NB (1): For Question 5.2, we only take into consideration nefopam+fentanyl group versus fentanyl group.Pain assessment:at rest and during movementat 12, 24, 36, 48 and 72 h after surgery using a visual analogue scale (VAS)NB (2): The evidence table reported only pain at H48 which was the only time point for which the difference was statistically significant.

Doses:nefopam group:nefopam 4 mg/hbolus: nefopam 2 mg (lock out ime = 15 min)

fentanyl group:fentanyl 20 µg/hbolus: fentanyl 10 µg (lock out ime = 15 min)

nefopam+fentanyl group:nefopam 1.86 mg/h + fentanyl 9.4 µg/hbolus: nefopam 0.93 mg + fentanyl 4.7 µg

(lock out ime = 15 min in every group)

DES

IRA

BLE

EFF

ECTS

How substantial are the desirable anticipated effects?○ Trivial● Small○ Moderate○ Large


This study has a "Non inferiority design"According to the study design, there was no inferiority of nefopam (adjunctive or not) compared to fentanyl (VAS non significant at H24 at rest and during movement).Note that fentanyl dose changes according to the group from 0 (no fentanyl, nefopam only) to 9.4 µg/h (nefopam+fentanyl group) to 20 µg/h (fentanyl only)VAS also non significant at H12, 24, 36, 48, 72 at rest, as well as during movement at H12, 24, 36, 72.The VAS difference between group is only significant during movement at

"large" desirable anticipated effects because nefopam did as well as fentanyl, a good point for nefopam if we would like to spare opioids.

H48:MD 0.34 higher (0.01 higher to 0.67 higher)i.e. VAS higher in nefopam+fentanyl group compared to fentanyl alone group

UN

DES

IRA

BLE

EFF

ECTS

How substantial are the undesirable anticipated effects?○ Large○ Moderate● Small○ Trivial


"small" undesirable anticipated effectsbeucause on the contrary, nausea was less frequent in the nefopam+fentanyl group (13%) compared to the fentanyl group (27%), p<0.01 taking into account the three groups (nefopam, nefopam+fentanyl, fentanyl).No significant difference between groups for: tachycardia, sweating, sedation, dyspnoa, prutirus, vomiting.(All adverse events were reported during the first 48 hours).No post-hoc tests comparing groups in pairs.

CER

TAIN

TY O

F EV

IDEN

CE What is the overall certainty of the

evidence of effects?○ Very low● Low○ Moderate○ High

○ No included studies

n=92 patients per group is a small group for John Centofanti.There was an important ROB:Randomization method not described, no intention to treat with 24/300 loss to follow-upReasons for non completion of the study:-duration of intubation > or = 48h: n=11 (it is not precised the timing of begining of the PCA with continuous mode, only that ot was "initiated postoperatively")-repeat surgery within 72h: n=5-reintubation in ICU: n=4-malfunctionning PCA device: n=4

"low" certainty of the evidence of effects because of:- a moderate to important size (almost 100!)- a double blinded designbut- only one study- with an important ROB

VA

LUES

Is there important uncertainty about or variability in how much people value the main outcomes?○ Important uncertainty or variability○ Possibly important uncertainty or variability○ Probably no important uncertainty or variability● No important uncertainty or variability○ No known undesirable outcomes

Except for nausea, other classical opioid related side effects were not decreased nor investigated (respiratory rate, ileus, delirium...).

"Possibly" important uncertainty or variability

BA

LAN

CE

OF

EFFE

CTS

Does the balance between desirable and undesirable effects favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison● Probably favors the intervention○ Favors the intervention


Globally no statistical difference (or small difference) of pain intensity but less nausea in nefopam+fentanyl group (intervention) compared to fentanyl alone group (comparison).

Also, there is at least a 50% reduction in fentanyl dose in the nefopam+fentanyl group (see dosing above, no difference in PCA volume between groups).

"the balance between desirable and undesirable effects favors the intervention"

RES

OU

RC

ES R

EQU

IRED

How large are the resource requirements (costs)?○ Large costs● Moderate costs○ Negligible costs and savings○ Moderate savings○ Large savings


Nefopam costs are very small in France (20 mg = $0.39; between $1.56 and $2.34 per day usually, 2016 prices in Montpellier Hospitals) but must probably depend on the country.Not available in the US and Canada.Countries where nefopam was available in 2010:BAHRAIN - TABLETS ONLYBELGIUMCHILEDOMINICAN REPUBLICEGYPTEQUATOR - TABLETS ONLYFRANCEGERMANY - TABLETS ONLYHAITIHONG KONG - TABLETS ONLYIRLANDLUXEMBOURGMALAISIAMALTAMAURITIUSMEXICONEW ZEALAND - TABLETS ONLYOMANPAKISTANPANANAQATARSALVADORSINGAPOUR - TABLETS ONLYSWITZERLANDTHE BARBADOS - TABLETS ONLYUNITED ARAB EMIRATESUNITED KINGDOM

"moderate costs"because this study used a PCA device but not the drug costs

CER

TAIN

TY O

F EV

IDEN

CE

OF

REQ

UIR

ED R

ESO

UR

CES

What is the certainty of the evidence of resource requirements (costs)?○ Very low● Low○ Moderate○ High

Cost and availability not known in every country See above.C

OST

EFF

ECTI

VEN

ESS

Does the cost-effectiveness of the intervention favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison● Probably favors the intervention○ Favors the intervention○ Varies○ No included studies

PCA was used in the two groups (no increased costs related to the PCA that was used in the two groups).Regarding the drug: nefopam might be less or more expensive than fentanyl (depends on local costs).Also to take into consideration, if nausea is less frequent, antiemetic drugs should be less required for rescue (this should decrease the averall cost of the drugs).NB: ramosetron was systematically given in both groups to reduce nausea.

cost-effectiveness "probably" favors the intervention

EQU

ITY

What would be the impact on health equity?○ Reduced○ Probably reduced● Probably no impact○ Probably increased○ Increased

AC

CEP

TAB

ILIT

Y Is the intervention acceptable to key stakeholders?○ No○ Probably no○ Probably yes○ Yes● Varies

the intervention is acceptable to key stakeholders "Probably YES"However, this requires patient's acceptation based on her/his preference:- some patients prefer avoiding opioids because of a past experience of nausea- some patients prefer, or do not prefer, using a PCA device

FEA

SIB

ILIT

YIs the intervention feasible to implement?○ No○ Probably no○ Probably yes○ Yes● Varies○ Don't know

Nefopam is widely available but not in all countries(please, see the countries where nefopam is available above).

"Varies"

Summary of judgementsJUDGEMENT IMPLICATIONS

PROBLEM No Probably no Probably yes Yes Varies Don't know

DESIRABLE EFFECTS Trivial Small Moderate Large Varies Don't know

UNDESIRABLE EFFECTS Large Moderate Small Trivial Varies Don't know

CERTAINTY OF EVIDENCE Very low Low Moderate High No included

studies

VALUES

Important uncertainty

or variability

Possibly important

uncertainty or variability

Probably no important


No important uncertainty or

variability

No known undesirable outcomes

BALANCE OF EFFECTSFavors the compariso

n

Probably favors the comparison

Does not favor either the

intervention or the comparison

Probably favors the

intervention

Favors the intervention Varies Don't know

RESOURCES REQUIRED Large costs Moderate costs Negligible costs and savings

Moderate savings Large savings Varies Don't know

CERTAINTY OF EVIDENCE OF REQUIRED RESOURCES

Very low Low Moderate High No included studies

JUDGEMENT IMPLICATIONS

COST EFFECTIVENESSFavors the compariso

n




Probably favors the

intervention

Favors the intervention Varies No included

studies

EQUITY Reduced Probably reduced

Probably no impact

Probably increased Increased Varies Don't know

ACCEPTABILITY No Probably no Probably yes Yes Varies Don't know

FEASIBILITY No Probably no Probably yes Yes Varies Don't know

Should adjunctive nefopam vs. no adjunctive nefopam be used for ICU pain management (5.2)?TYPE OF RECOMMENDATION

Strong recommendation against the intervention

Conditional recommendation against

the intervention

Conditional recommendation for either

the intervention or the comparison

Conditional recommendation for the

intervention

Strong recommendation for the intervention

○ ○ ○ ● ○

RECOMMENDATION We suggest that adjunctive nefopam (if available) can be used to decrease opioid dose and nausea while treating pain in ICU patients (at risk of opioid side effects) (+2C).no recommendationnausea considered in previous recommendation comparing nefopam vs fentanyl groups; no clinical outcomes difference, reduction in side effects

JUSTIFICATION Overall justificationOnly one moderate size, moderate quality (8% of excluded patients, no intent to treat analysis) doube blind RCT in 184 postcardiac SICU patients able to use a PCA device.-> Level B of evidence because the RCT is downgraded.Detailed justificationProblemVery selected ICU population. However, generalization of the results to another kind of population is likely because pain is frequent in postcardiac SICU patients.FeasibilityPatients able to use a PCA device. These promising results need to be also found in patients unable to use a PCA device (non communicant

patients). However, 62% of nefopam +fentanyl infusion was provided continuously (not by PCA) at H12, and 77% at H48.

SUBGROUP CONSIDERATIONS

IMPLEMENTATION CONSIDERATIONS

Availability of nefopam in some countries, availability of a PCA device in some ICUs.

MONITORING AND EVALUATION

Analgesic related side effects or outcomes should be assessed more largely including:-delirium-ileus-duration of MV weaning-LOS in ICU and hospital

RESEARCH PRIORITIES This is top priority to investigate adjunctive analgesics that could decrease opioids dose to treat pain in ICU patients.These promising results require that this study be replicated:- in medical ICU and noncardiac SICU patients- non communicant patients- using a more common mode of drug administration in ICU patients who are sometimes non communicant (a continuous infusion with boli and/or an intermittent administration, but provided by the nurse based on behavioral pain tools, instead of a PCA device needing the participation of the patient)

COMMENTS DURING ELECTRONIC VOTING BY ENTIRE PANEL

Adding nefopam did not appear to change pain over time; do lower fentanyl doses translate into clinical benefit?Nausea (side effect) has a risk of bias in this specific population; Inexperience using this medication mitigates enthusiasmRephrazing to consider nefopam alone vs opioid alone: should adjunctive nefopam or nefopam alone (vs. an opioid alone) be used…?Very low quality of evidence, how can a decision be stated?

Question: Nefopam compared to Opioids for ICU Patients (5.1) Setting: Post-operative Cardiac Surgery Patients Bibliography: Kim K., Kim WJ., Choi DK., Lee YK, Choi I., Sim J. The analgesic efficacy and safety of nefopam in patient-controlled analgesia after cardiac surgery: A randomized, double-blind, prospective study. Journal of International Medical Research: 42(3), 684-692.

Quality assessment № of patients Effect


s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other considerations Nefopam Opioids

Relative

(95% CI)

Absolute(95% CI)

Pain at rest at 24 hours (follow up: mean 24 Hours; assessed with: Visual Analogue Scale; Scale from: 0 to 10)

1 randomised trials

serious a not serious not serious b serious c none 92 92 - mean 0.23 mean higher(0.21 lower to 0.68 higher)

⨁⨁◯◯LOW

IMPORTANT

Pain with movement at 24 hours (follow up: mean 24 hours; assessed with: Visual Analogue Scale; Scale from: 0 to 10)



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other considerations Nefopam Opioids

Relative

(95% CI)

Absolute(95% CI)

1 randomised trials

serious a not serious not serious b serious c none 92 92 - mean 0.14 mean higher(0.28 lower to 0.56 higher)

⨁⨁◯◯LOW

IMPORTANT

Pain at rest at 48 hours (assessed with: Visual Analogue Scale)

1 randomised trials

serious a not serious not serious b serious c none 92 92 - MD 0.25 higher

(0.553 lower to 0.053 higher)

⨁⨁◯◯LOW

IMPORTANT

Pain with movement at 48 hours (follow up: mean 24 hours; assessed with: Visual Analogue Scale; Scale from: 0 to 10)

1 randomised trials

serious a not serious not serious b serious c,d none 92 92 - MD 0.41 higher

(0.081 higher to 0.739 higher)

⨁⨁◯◯LOW

IMPORTANT

CI: Confidence interval; MD: Mean differenceExplanationsa. Randomization method not described, no intention to treat with 24/300 loss to follow-up, blinding described b. Only included CVICU patients in this study, however analgesic efficacy may be broadly applicable to all ICU patients c. Large confidence intervals, cannot determine if value of nefopam independently d. Confidence Intervals do not cross zero (no effect) - however small sample size hard to draw conclusion on nefopam effects

QuestionShould Nefopam vs. Opioids be used for ICU Patients (5.1)?

POPULATION: ICU Patients (5.1) BACKGROUND:

INTERVENTION:

Nefopam

COMPARISON: Opioids

MAIN OUTCOMES:

Pain at rest at 24 hours; Pain with movement at 24 hours; Pain at rest at 48 hours; Pain with movement at 48 hours;

SETTING: Post-operative Cardiac Surgery Patients

PERSPECTIVE:

AssessmentJUDGEMENT RESEARCH EVIDENCE ADDITIONAL CONSIDERATIONS

PRO

BLE

M

Is the problem a priority?○ No○ Probably no○ Probably yes● Yes


It is top priority to investigate the effectiveness/side effects of Non opioid analgesics compared to opioids in order to decrease opioids related side effects that are especially at risk in ICU patients:- respiratory depression- neurological impairment (decreased vigilance, delirium)- ileus, nausea vomiting, etc...Kim et al. 2014[1]n=276 patients scheduled to undergo cardiac surgery were randomly assigned between threePCA groups (92 patients per group): nefopam, fentanyl or nefopam + fentanyl.Each drug was given continuously + in PCA mode.NB (1): For Question 5.1, we only take into consideration nefopam group versus fentanyl group.Pain assessment:at rest and during movementat 12, 24, 36, 48 and 72 h after surgery using a visual analogue scale (VAS)

NB (2): The evidence table reported only pain at rest and with movement at H24 and H48.

Doses:nefopam group:nefopam 4 mg/hbolus: nefopam 2 mg (lock out ime = 15 min)fentanyl group:fentanyl 20 µg/hbolus: fentanyl 10 µg (lock out ime = 15 min)nefopam+fentanyl group:nefopam 1.86 mg/h + fentanyl 9.4 µg/hbolus: nefopam 0.93 mg + fentanyl 4.7 µg(lock out ime = 15 min for every group)

DES

IRA

BLE

EFF

ECTS



This study has a "Non inferiority design"According to the study design, there was no inferiority of nefopam compared to fentanyl(VAS non significant at H24 at rest and during movement).VAS also non significant at H12, 24, 36, 48, 72 at rest, as well as during movement at H12, 24, 36, 72.The VAS difference between group is only significant during movement at H48:MD 0.41 higher (0.081 higher to 0.739 higher)i.e. VAS higher in nefopam group compared to fentanyl group

"large" desirable anticipated effects because nefopam did as well as fentanyl, a good point for nefopam if we would like to spare opioids.Is nefopan better than opioids? Different approach from the one we used in previous questions. Small effect.-agrees; large would mean that nefopam is better than opioids; small-agrees with smallGroup consensus for "small effects"

UN

DES

IRA

BLE

EFF

ECTS

How substantial are the undesirable anticipated effects?○ Large○ Moderate● Small○ Trivial


"small" undesirable anticipated effects because on the contrary,nausea was less frequent in the nefopam group (11%) compared to the fentanyl group (27%), p<0.05.-No significant difference between groups for:tachycardia, sweating, sedation, dyspnea, prutirus, vomiting. (All adverse events were reported during the first 48 hours).-could be moved to moderate conisdeing that they were fewer side effects with nefopam; are their adverse effects of nefopam that could be detrimental?-delirium (but not reported)-curious to know how they found out about nausea in MV patients: moderate or small would be fine-small because the difference is not large between the 2 groupsnefopam has small undesirable effects compare to opioids

CER

TAIN

TY O

F EV

IDEN

CE What is the overall certainty of

the evidence of effects?○ Very low● Low○ Moderate○ High


n=92 patients per group is a small group for John Centofanti. There was an important ROB:Randomization method not described, no intention to treat with 24/300 loss to follow-up Reasons for non completion of the study:-duration of intubation > or = 48h: n=11 (it is not precised the timing of begining of the PCA with continuous mode, only that ot was "initiated postoperatively")-repeat surgery within 72h: n=5-reintubation in ICU: n=4-malfunctionning PCA device: n=4

"moderate" certainty of the evidence of effects because of:- a moderate to important size (almost 100!)- a double blinded designbut- only one study- with an important ROB- imprecision noted (CI) + ROB assessment- Low because only one study; power analysis missing; non-significant resultsGroup: keep "low"

VA

LUES

Is there important uncertainty about or variability in how much people value the main outcomes?○ Important uncertainty or variability○ Possibly important uncertainty or variability○ Probably no important uncertainty or variability● No important uncertainty or variability

○ No known undesirable outcomes

Except for nausea, other classical opioid related side effects were not decreased nor investigated (respiratory rate, ileus, delirium...).

"Possibly" important uncertainty or variabilityIn previous questions, we referred to pain scores as the main outcome. VAS is a valid and accepted measure.

BA

LAN

CE

OF

EFFE

CTS

Does the balance between desirable and undesirable effects favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison● Probably favors the intervention○ Favors the intervention○ Varies○ Don't know

Globaly no statistical difference (or small difference) of pain intensity but less nausea in nefopam group (intervention) compared to fentanyl group (comparison).

Gerald: "the balance between desirable and undesirable effects favors the intervention" because:-Probably favors because the only difference was for nausea; only one study too- balancing between nefopam and opioids-also consider the possibility to reduce exposure to opioids

RES

OU

RC

ES R

EQU

IRED

How large are the resource requirements (costs)?○ Large costs● Moderate costs○ Negligible costs and savings○ Moderate savings○ Large savings


Nefopam costs are very small in France (20 mg = $0.39; between $1.56 and $2.34 per day usually, 2016 prices in Montpellier Hospitals) but must probably depend on the country.

Not available in the US and Canada.Countries where nefopam was available in 2010:

See above for list of countries where marketed

"moderate costs"because this study used a PCA device but not regarding the drug costs

CER

TAIN

TY O

F EV

IDEN

CE

OF

REQ

UIR

ED R

ESO

UR

CES

What is the certainty of the evidence of resource requirements (costs)?○ Very low● Low○ Moderate○ Higg○ No included studies

See above.C

OST

EFF

ECTI

VEN

ESS

Does the cost-effectiveness of the intervention favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison● Probably favors the intervention○ Favors the intervention○ Varies○ No included studies

PCA was used in the two groups (no increased costs related to the PCA that was used in the two groups).Regarding the drug: nefopam might be less or more expensive than fentanyl (depends on local costs).Also to take into consideration, if nausea is less frequent, antiemetic drugs should be less required for rescue (this should decrease the averall cost of the drugs).NB: ramostron was systematically given in both groups to reduce nausea.

cost-effectiveness "probably" favors the intervention

EQU

ITY

What would be the impact on health equity?○ Reduced○ Probably reduced● Probably no impact○ Probably increased○ Increased○ Varies○ Don't know

AC

CEP

TAB

ILIT

Y

Is the intervention acceptable to key stakeholders?○ No○ Probably no○ Probably yes○ Yes● Varies○ Don't know

intervention is acceptable to key stakeholders "Probably YES"because this requires patient's acceptation based on her/his preference:- some patients prefer avoiding opioids because of a past experience of nausea for example- some patients prefer, or do not prefer, using a PCA deviceIt varies because of the availability of the drug and the use of PCA.Varies because also depends on patient's preference.

FEA

SIB

ILIT

YIs the intervention feasible to implement?○ No○ Probably no○ Probably yes○ Yes● Varies○ Don't know

Nefopam is widely available but not in all countries (Canada and the US for example... but these guidelines are international :)

"varies"






studies

VALUESImportant


Possibly important





variability


BALANCE OF EFFECTS

Favors the comparison




Probably favors the

intervention


RESOURCES REQUIRED Large costs Moderate costs Negligible costs

and savingsModerate savings Large savings Varies Don't know



COST EFFECTIVENESS




Probably favors the

Favors the intervention

Varies No included studies


intervention or the comparison intervention


Probably no impact




Should Nefopam vs. Opioids be used for ICU Patients (5.1)?TYPE OF RECOMMENDATION Strong

recommendation against the intervention

Conditional recommendation

against the intervention

Conditional recommendation for

either the intervention or the comparison


the intervention

Strong recommendation for

the intervention

○ ○ ● ○ ○

RECOMMENDATION We suggest that nefopam (if available) can be an alternative to opioids to treat pain in ICU patients, especially in case of opioids contraindication (+2C).Change "might" with "can"; keep it general and specify in the justification the sub-group considerations.

JUSTIFICATION Overall justificationOnly one moderate size, moderate quality (8% of excluded patients, no intent to treat analysis) doube blind RCT in 184 postcardiac SICU patients able to use a PCA device.-> Level B of evidence because the RCT is downgraded.Detailed justificationProblemVery selected ICU population. However, generalization of the results to another kind of population is likely because pain is frequent in postcardiac SICU patients.FeasibilityPatients were able to use a PCA device. These promising results need to be also found in patients unable to use a PCA device (non communicant patients). However, 63% of nefopam infusion was provided continuously (not by PCA) at H12, and 83% at H48.

SUBGROUP CONSIDERATIONS Cardiac surgery patients

IMPLEMENTATION CONSIDERATIONS Availability of nefopam in some countries, availability of a PCA device in some ICUs.

MONITORING AND EVALUATION Analgesic related side effects or outcomes should be assessed more largely including:

-delirium-ileus-duration of MV weaning-LOS in ICU and hospital

RESEARCH PRIORITIES This is top priority to investigate alternative analgesics to opioids to treat pain in ICU patients.These promising results require that this study be replicated:- in medical ICU and noncardiac SICU patients/ - non-communicative patients- using a more common mode of drug administration in ICU patients who are challenged when communicating (e.g. a continuous infusion with boluses and/or an intermittent administration, provided by the nurse based on behavioral pain tools, instead of a PCA device needing the participation of the patient)


Nefopam of uncertain benefit for pain given no change at most time points- despite non-inferiority trial. Although fentanyl dose lower, how did that translate into a clinical benefit? Nausea finding has risk of bias (specific patient population); Limited applicability to US.Rephrazing to consider nefopam alone vs opioid alone: adjunctive nefopam or nefopam alone (vs. an opioid alone) be used…? Should we consider this opiate-sparing strategy? Very low quality of evidence, how can a decision be stated? Inexperience using this medication mitigates enthusiasm

Question: Adjunctive paracetamol compared to no adjunctive paracetamol for ICU pain management (5.2) Setting: Intensive Care Unit


Quality Importance

№ of studie

s

Study design

Risk of

bias

Inconsistency

Indirectness

Imprecision

Other conside-rations

adjunctive paracetamol

no adjunctive

paracetamol

Relative(95% CI)

Absolute(95% CI)

VAS Score at 24 hours postoperatively (in cm)

2 randomised trials

serious a

not serious not serious b not serious none 76 77 - MD 0.46 lower(0.69 lower to

0.23 lower)

⨁⨁⨁◯MODERAT

E

CRITICAL

Mean BPS Pain Scores until patient extubated

1 randomised trials

very serious c

not serious not serious d serious e none 20 20 - MD 1.98 lower(2.98 lower to

0.98 lower)


CRITICAL

Pain Score at extubation

1 randomised trials

very serious c

not serious not serious d serious e none 20 20 - MD 1.1 lower(1.73 lower to

0.47 lower)


CRITICAL

Time to extubation (minutes)


Quality Importance

№ of studie

s

Study design

Risk of

bias

Inconsistency

Indirectness

Imprecision

Other conside-rations


no adjunctive

paracetamol

Relative(95% CI)

Absolute(95% CI)

1 randomised trials

very serious c

not serious not serious d serious e none 20 20 - MD 140.2 lower

(192.7 lower to 87.7 lower)


CRITICAL

Rescue Doses of Morphine

1 randomised trials

not serious

not serious not serious f serious g none 8/56 (14.3%) 14/57 (24.6%)

OR 0.51(0.20 to

1.34)

103 fewer per 1,000

(from 58 more to 184 fewer)

⨁⨁⨁◯MODERAT

E

CRITICAL

Opioid Consumption (in Morphine equivalents)

2 randomised trials

serious a

not serious not serious b not serious none 76 77 - MD 4.54 lower(6.6 lower to 2.47 lower)

⨁⨁⨁◯MODERAT

E

CRITICAL

CI: Confidence interval; MD: Mean difference; OR: Odds ratioExplanations a. 1 study has low ROB, however the 2nd study was high ROB due to lack of blinding, no intention to treat and unclear randomization process b. Includes both postoperative cardiac surgery and abdominal surgery patients c. No blinding, unclear randomization process, no intention to treat d. Includes only abdominal surgery ICU pts. Lowered due to very small sample size f. Includes only cardiac surgery ICU patients g.OR CI cannot rule out harm at upper limit QuestionShould adjunctive paracetamol vs. no adjunctive paracetamol be used for ICU pain management (5.2)?


INTERVENTION:


COMPARISON: no adjunctive paracetamol

MAIN OUTCOMES:

VAS Score at 24 hours postoperatively (in cm); Mean BPS Pain Scores until patient extubated; Pain Score at extubation; Time to extubation (minutes); Rescue Doses of Morphine; Opioid Consumption (in Morphine equivalents);

SETTING: Intensive Care Unit

PERSPECTIVE:


PRO

BLE

MIs the problem a priority?○ No○ Probably no○ Probably yes● Yes


It is top priority to investigate the effectiveness/side effects of Adjunctive Non opioid analgesics to opioids compared to opioids alone in order to decrease opioids related side effects that are especially at risk in ICU patients:- respiratory depression- neurological impairment (decreased vigilance, delirium)- ileus, nausea vomiting, etc...Two single center RCT in ICU patients after a scheduled surgery- Cattabriga 2007 [2]: cardiac surgery, n=113, low ROB- Memis 2010 [3]: abdominal surgery, n=40, unblinded, high ROBIntervention:Adjunct IV paracetamol 1g/6hrs in both studies- Cattabriga [2]: for 72 hours, opioid= tramadol + rescue morphine- Memis [3]: for 24 hours, opioid= mepiridine + rescue mepiridinePain measurement:VAS in Cattabriga [2] (/6h till H72, at rest and during deep breath),mean VAS after extubation in Memis (at H24 postextubation);mean BPS before and at extubation in Memis

DES

IRA

BLE

EFF

ECTS How substantial are the desirable

anticipated effects?○ Trivial○ Small● Moderate○ Large


Mean BPS before extubation (Memis) [3] : MD 3 fewer (3 fewer to 1 fewer) (5.7 versus 3.7)Mean BPS at extubation (Memis) [3]: MD 1 fewer (1.7 fewer to 0.5 fewer) (3.6 versus 2.5 -> it is weird to show a result of mean BPS < 3??)Forrest plot for VAS (cm) supposed to be at rest in both studies = -0.5 [-0.7 to -0.2]In fact, Cattabriga has the highest weight (highest reduction in VAS) but no difference in opioid consumption. Inversely, there was a less important reduction of VAS in Memis [3]but a significant reduction in opioid consumption.I think is mathematical/logical. Adjunct paracetamol has "additional" analgesic effect.NB: mean VAS are low in all groups in both studies (means or medians < 3)Forrest plot for opioid consumption (mg morphine equivalents, within 24h for Memis; 72h for Cattabriga [2]) = -4.5 [-6.6 to -2.5]-was the forrest plot mean the consumption per 24h or was it an "overall" forest plot taking into account the consumption within 24h in Memis and in 72h in Catabriga?Rescue dose of morphine (Cattabriga [2]) : OR 0.51 (0.20 to 1.34)Other outcomes:Time to extubation in minutes (Memis[3]): MD 140 fewer (193 fewer to 88 fewer)Sedation score (1 to 5 points scale) significantly lower in Memis (= less sedated)Nausea lower in Memis[3], Nausea NS in Cattabriga [2]ICU LOS, respiratory depression, reintubation, pruritus: NS in MemisRespiratory rate significantly lower in paracetamol group at H12 only in Cattabriga

"moderate" desirable anticipated effectsbecause mean VAS are low in all groups in both studies but there is a significant reduction of VAS and opioid consumption.

UN

DES

IRA

BLE

EFF

ECTS

How substantial are the undesirable anticipated effects?○ Large○ Moderate○ Small● Trivial○ Varies○ Don't know

"small" undesirable anticipated effectsbecause no adverse outcome was reported, on the contrary: decreased duration of MV and nausea in one studyParacetamol is a safe drug if dose is appropriately given (according to patient's weight, renal insuficiency). Otherwise, there is a risk of hepatitis and renal failure. Not assessed in these studies.Question: No adverse events reported, what about vital parameters (e.g., blood pressure)? in association with dosing smalldon't knowno differences in BP between groups in Cattabriga [2] and Memis [3]trivial (of little value or importance)Trivial based on information we have

CER

TAIN

TY O

F EV

IDEN

CE

What is the overall certainty of the evidence of effects?○ Very low○ Low● Moderate○ High○ No included studies

Two RCTs but single center, one double blinded with a low ROB (Cattabriga) [2], one unblinded with a high ROB, that was also a very small size study (Memis) [3].

"moderate" certainty of the evidenceV

ALU

ES


"Probably no important" uncertainty or variabilityCeline: No important uncertainty when we have solid measures of outcomes (pain scores, opioid doses)

BA

LAN

CE

OF

EFFE

CTS

Does the balance between desirable and undesirable effects favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison● Probably favors the intervention○ Favors the intervention○ Varies○ Don't know

Probably favors the interventionbecause significant reduction in VAS and opioid consumption and probably lower rate of adverse outcomes (in the lowest quality study)

RES

OU

RC

ES

REQ

UIR

ED

How large are the resource requirements (costs)?○ Large costs○ Moderate costs○ Negligible costs and savings○ Moderate savings○ Large savings● Varies○ Don't know

IV Paracetamol in not expensive in France (Price in Montpellier hospitals in 2016): $0.69 for 1 gramm IV ($2.76 for 4 gramms a day).Is it the same case in the US/Canada?Not available in Canada

Negligible costs and savingsbecause IV paracetamol is not expensive and because it is associated with savings related to reduced opioid consumption.costs will vary based on the country

CER

TAIN

TY O

F EV

IDEN

CE

OF

REQ

UIR

ED R

ESO

UR

CES

What is the certainty of the evidence of resource requirements (costs)?○ Very low○ Low○ Moderate○ High● No included studies

Not addressed in the 2 RCTsC

OST

EFF

ECTI

VEN

ESS

Does the cost-effectiveness of the intervention favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison○ Probably favors the intervention○ Favors the intervention○ Varies● No included studies

Not addressed in the 2 RCTs Probably favors the intervention

EQU

ITY


AC

CEP

TAB

ILIT

Y Is the intervention acceptable to key stakeholders?○ No○ Probably no● Probably yes○ Yes○ Varies○ Don't know

Considering all routes of administration "Probably YES"

FEA

SIB

ILIT

YIs the intervention feasible to implement?○ No○ Probably no○ Probably yes● Yes○ Varies○ Don't know

the intervention is feasible to implement-> "YES"because systematic administration of paracetamol is easy to implement






studies

VALUESImportant


Possibly important





variability


BALANCE OF EFFECTS





Probably favors the

intervention







COST EFFECTIVENESS





Probably favors the intervention


studies


Probably no impact




Should adjunctive paracetamol vs. no adjunctive paracetamol be used for ICU pain management (5.2)?TYPE OF RECOMMENDATION


Conditional recommendation against the

intervention




intervention


○ ○ ○ ● ○

RECOMMENDATION We suggest that adjunctive paracetamol might be used to decrease pain intensity and opioid consumption in ICU patients (+2C).Remove "IV" because not part of the question; keep the recommendation generalRCTs done with IV paracetamolIV comparable to po route of administration in previous studies

JUSTIFICATION Overall justificationTwo RCT but single centers, one of very low quality (unblinded, high ROB, small size).-> level of evidence C because the study that showed less undesirable effects has a low quality (unblinded, hogh ROB), it was downgraded to C. The high quality study that showed the highest reduction in VAS showed also low VAS in both groups and no additional effects.Detailed justificationProblemSurgical ICU population (one cardiac, one abdominal, both scheduled). However, generalization of the results to another kind of population is likely because pain is frequent in SICU patients as in medical ICU patients.


Cardiac and abdominal surgery patients


NoneIV route not available in Canada, and expensive in US - but other routes can be used


Pain assessmentAnalgesic related side effects or outcomes should be assessed more largely including:-delirium-ileus-duration of MV weaning-LOS in ICU and hospitalParacetamol related side effects should be assessed more specifically: increased serum liver enzymes (acute hepatitis), decreased liver and kidney function

RESEARCH PRIORITIES

This is top priority to investigate analgesics that can decrease the use or dose of opioids to treat pain in ICU patients. These promising results require that these studies be replicated:- in medical ICU; other surgeries and trauma- non-verbal patients (only one small size study assessed the BPS before extubationOther routes of paracetamol administration to be investigated


with very low quality of evidence should the recommendation be as stated? or re-phrased to indicate you cannot make a recommendation?

Question: Adjunctive ketamine compared to no adjunctive ketamine for ICU analgesic management (5.2) Setting: Intensive Care Unit Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other consideration

s

adjunctive

ketamine

no adjunctiv

e ketamine

Relative(95% CI)

Absolute(95% CI)

Cumulative morphine consumption (follow up: mean 48 hours; assessed with: dose (mg))

1 randomised trials

serious a not serious serious b not serious none 41 52 - MD 22 mg fewer

(14 fewer to 30 fewer)

⨁⨁◯◯LOW

IMPORTANT

VAS score at rest (follow up: mean 48 hours; assessed with: VAS score (mm); Scale from: 0 to 100)

1 randomised trials

serious a not serious serious b serious c none 41 52 - MD 3 mm lower



CRITICAL

VAS score with movement (follow up: mean 48 hours; assessed with: VAS Score (mm); Scale from: 0 to 100)

1 randomised trials

serious a not serious serious b serious c none 41 52 - MD 4 mm lower



CRITICAL

CI: Confidence interval; MD: Mean differenceExplanationsa. Unclear randomization process and location concealment, no intention to treat analysis, b. Only postoperative major abdominal surgery patients included c. Unclear benefit, small sample size

QuestionShould adjunctive ketamine vs. no adjunctive ketamine be used for ICU analgesic management (5.2)?

POPULATION: ICU analgesic management (5.2) BACKGROUND:

INTERVENTION:

adjunctive ketamine

COMPARISON: no adjunctive ketamine

MAIN OUTCOMES:

Cumulative morphine consumption; VAS score at rest; VAS score with movement;


PERSPECTIVE:


PRO

BLE

M

Is the problem a priority?○ No○ Probably no○ Probably yes● Yes○ Varies○ Don't know

It is top priority to investigate the effectiveness/side effects of Adjunctive Non opioid analgesics to opioids compared to opioids alone in order to decrease opioids related side effects that are especially at risk in ICU patients:- respiratory depression- neurological impairment (decreased vigilance, delirium)- ileus, nausea vomiting, etc...One single center double-blinded RCT in 93 postscheduled abdominal surgery ICU patients, high ROB (unclear randomization process and location concealment, no intention to treat analysis despite 8 patients excluded after enrollment) -> Guillou 2003 [4]two parallel groups:1) morphine PCA + ketamine initial loading charge + kétamine continuous infusion2) morphine PCA + placebo loading charge + placebo continuous infusionNB: intervention began after patients' awakening in the ICUPain intensity at rest at H48 (VAS in mm; 0-100):MD 3 mm lower (11 mm lower to 5 mm higher) -> NSPain intensity at mobilization at H48 (VAS in mm; 0-100):MD 4 mm lower (15 mm lower to 7 mm higher) -> NSNote than mean VAS were small in both groupsMorphine consumption at H48 (mg)MD 22 mg fewer (30 fewer to 14 fewer)Side effects were NS:nausea, confusion, hallucinations, hypoventilation, pruritus;Ramsay scale similar in the two groups

DES

IRA

BLE

EFF

ECTS How substantial are the

desirable anticipated effects?○ Trivial● Small○ Moderate○ Large○ Varies○ Don't know

"Small desirabe anticipated effects"because in favour to the intervention (adjunctive ketamine) group:- Small difference in VAS at every time points (p<0.05 only at H16, H44 at test; at H16, H20, H40 at mobilization)- Significant but small difference in cumulative morphine consumption at each time point (every 4-hours from H4 to H48): 22mg at H48 = a mean of 11 mg/24h

UN

DES

IRA

BLE

EF

FEC

TS

How substantial are the undesirable anticipated effects?○ Large○ Moderate○ Small○ Trivial○ Varies● Don't know

"don't know undesirabe anticipated effects"because adverse effects that were measured were not frequent but the method for assessing confusion is not described, other adverse effects related to ketamine were not reported, such as nightmare, anxiety, and delirium. Some of the adverse effects related to opioids were not reported either: ileus

CER

TAIN

TY O

F EV

IDEN

CE

What is the overall certainty of the evidence of effects?● Very low○ Low○ Moderate○ High○ No included studies

Only one single small size RCT with a high ROB: unclear randomization process and location concealment, no intention to treat analysis despite 8 patients excluded after enrollment

low" certainty of the evidence fo the effects

Change with "very low" for consistency with evidence table

VA

LUES


Possibly important uncertainty or variability" because the desirable effects are small and the undesirable effects seem to be small but under-reported: delirium should be measured with valid tools; psychodysleptic effects related to ketamine should be reportedNo important uncertainty about main outcomes (pain scores, opioids)

BA

LAN

CE

OF

EFFE

CTS

Does the balance between desirable and undesirable effects favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison○ Probably favors the intervention○ Favors the intervention○ Varies● Don't know

"don't know if the balance between desirable and undesirable effects favor the intervention or the comparison"because small desirable effects but don't know about the undesirable effects

RES

OU

RC

ES R

EQU

IRED How large are the resource

requirements (costs)?○ Large costs● Moderate costs○ Negligible costs and savings○ Moderate savings○ Large savings○ Varies○ Don't know

Ketamine costs are small. "moderate costs"because this study used a PCA device

CER

TAIN

TY O

F EV

IDEN

CE

OF

REQ

UIR

ED

What is the certainty of the evidence of resource requirements (costs)?○ Very low○ Low○ Moderate○ High● No included studies

See above

CO

ST E

FFEC

TIV

ENES

SDoes the cost-effectiveness of the intervention favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison○ Probably favors the intervention○ Favors the intervention○ Varies● No included studies

Not addressed in this study Cost-effectiveness "Does not favor either the intervention or the comparison"because the PCA device is used in both groups, the ketamine costs are small,the desirable effects are small, the undesirable effects are unknown

EQU

ITY


AC

CEP

TAB

ILIT

Y

Is the intervention acceptable to key stakeholders?○ No○ Probably no● Probably yes○ Yes○ Varies○ Don't know

Is the intervention acceptable to key stakeholders? "-> "Probably YES"However, this requires patient's acceptation based on her/his preference:- some patients prefer, or do not prefer, using a PCA device-for some patients, ketamine may not be acceptable

FEA

SIB

ILIT

Y

Is the intervention feasible to implement?○ No○ Probably no● Probably yes○ Yes○ Varies○ Don't know

implementation feasibility "varies"because it depends on the PCA availability; probably yes because it's feasible to implement ketamine






studies

VALUESImportant


Possibly important





variability


BALANCE OF EFFECTS











COST EFFECTIVENESS







studies



Probably no impact




Should adjunctive ketamine vs. no adjunctive ketamine be used for ICU analgesic management (5.2)?TYPE OF RECOMMENDATION



intervention




intervention


○ ○ ● ○ ○

RECOMMENDATION We suggest that either adjunctive ketamine along with morphine or morphine alone be used to treat postoperative pain in ICU patients (+2C).Remove PCA. Consider meta-analysis reports in our recommendation?utility in ICU is understudied

JUSTIFICATION Only one single small size low quality RCT in ICU patients.-> level of evidence downgraded to CAlso to take into consideration the recent reviews regarding ketamine to prevent/treat postoperative pain ("Ketamine added to morphine or hydromorphone patient-controlled analgesia for acute postoperative pain in adults: a systematic review and meta-analysis of randomized trials, by Wang et al. Canadian Journal of anesthesia 2016" and "Benefit and harm of adding ketamine to an opioid in a patient-controlled analgesia device for the control of postoperative pain: systematic review and meta-analyses of randomized controlled trials with trial sequential analyses, by Assouline et al. Pain december 2016). Authors’ conclusions are pretty different: 1) Wang, CJA/JCA 2016 [5]Adding ketamine to morphine/hydromorphone PCA provides a small improvement in postoperative analgesia while reducing opioid requirements. Adjunctive ketamine also reduces postoperative nausea and vomiting without a detected increase in other adverse effects; however, adverse events were probably underreported. 2) Assouline, Pain 2016 [6]Trial sequential analyses confirmed the significant benefit of ketamine on pain intensity, cumulative morphine consumption, and postoperative nausea and vomiting and its inability to double the risk of hallucination. The available data did not allow us to make a conclusion on respiratory adverse events or to establish dose-responsiveness.


abdominal surgery


Availability of a PCA device in some ICUs.


Analgesic related adverse effects or outcomes should be assessed more largely including:-delirium assessed with validated tools for ICU patients

-psychodysleptic effects related to ketamine (hallucinations, delusion, nightmare, anxiety...)-ileus-duration of MV weaning-LOS in ICU and hospital

RESEARCH PRIORITIES

This is top priority to investigate analgesics that can decrease the use or dose of opioids to treat pain in ICU patients. There is only one negative study in postoperative abdominal surgery ICU patients. Recent updated reviews (2016) of trials investigating the effect of adjunctive ketamine to opioids administered with PCA to treat postoperative pain show a significant reduction in pain intensity, opioid consumption, nausea/vomiting. Other side effects like delirium could have been under reported. Future studies are required:- in medical ICU patients experiencing severe pain at rest (pancreatitis for example)- in medical and surgical ICU patients, especially non communicant patients at the early stage of sedation-analgesia, via a continuous infusion of ketamine without the need of a PCA-patient's preference and satisfaction with ketamine-other RCTs necessary in an ICU context


Where is the level of evidence? Broad recommendation for post-op. pain in one study with high ROB; no guidance to clinicians on when it might be (in)appropriate.Primary ketamine infusions for sedation and analgesia? need to be clear this is not a first line therapy…Our "suggest" recommendation could change clinician's practice given growing interest in ketamine. Caution: high risk of bias in the single study & the most important psychiatric side effects of ketamine is unreported. Need clinical benefits beyond reduced opioid consumption, psychiatric side effect analysis, good sample and low ROB.

Question: Adjunct neuropathic agent compared to no adjunct neuropathic drug for ICU pain management (5.2) Setting: Medical and Surgical ICUs Bibliography:


Quality Importance

№ of studie

s

Study design

Risk of

bias

Inconsistency

Indirectness

Imprecision

Other consideration

s

adjunct neuropathi

c agent

no adjunct neuropathi

c drug

Relative

(95% CI)

Absolute(95% CI)

# of patients with VRS >=2/4, 10 hours post-extubation

1 randomised trials

serious a

not serious not serious b serious c none 3/29 (10.3%)

12/31 (38.7%)

RR 0.27(0.08 to

0.85)

283 fewer per 1,000

(from 58 fewer to 356 fewer)

⨁⨁◯◯LOW

CRITICAL

Opioid consumption in first 24 hours (morphine equivalent)

4 randomised trials

serious d

not serious not serious not serious none 97 97 - MD 13.54 lower(14.57 lower to

12.5 lower)

⨁⨁⨁◯MODER

ATE

CRITICAL

Time to Extubation (hours)


Quality Importance

№ of studie

s

Study design

Risk of

bias

Inconsistency

Indirectness

Imprecision

Other consideration

s

adjunct neuropathi

c agent

no adjunct neuropathi

c drug

Relative

(95% CI)

Absolute(95% CI)

2 randomised trials

serious a

not serious not serious b serious e none 49 51 - MD 0.36 higher(0.7 lower to 1.43 higher)

⨁⨁◯◯LOW

CRITICAL

ICU LOS

2 randomised trials

serious a

not serious not serious b serious f none 49 51 - MD 0.04 lower(0.46 lower to 0.38 higher)

⨁⨁◯◯LOW

CRITICAL

NRS Score Day 4 after Neuropathic Agent initiated

2 randomised trials

not serious

not serious not serious g not serious none 42 30 - MD 3.44 lower(3.9 lower to 2.98

lower)

⨁⨁⨁⨁HIGH

CRITICAL

Breakthrough pain in first 48hrs (average #/pt)

1 randomised trials

not serious

not serious not serious serious e none 20 20 - MD 1.45 lower(2.13 lower to

0.77 lower)

⨁⨁⨁◯MODER

ATE

CRITICAL

CI: Confidence interval; RR: Risk ratio; MD: Mean differenceExplanations a. Unclear randomization process, no intention to treat with 10 patients excluded from analysis post-randomization b. Only includes cardiac surgery patients c. Low event-rate d. Unclear randomization, no intention to treat e. Small sample size f. Unclear if any benefit or harm g. Only includes patients with Guillain Barre

QuestionShould adjunct neuropathic agent vs. no adjunct neuropathic drug be used for ICU pain management (5.2)?


INTERVENTION:

adjunct neuropathic agent

COMPARISON: no adjunct neuropathic drug

MAIN OUTCOMES:

# of patients with VRS >=2/4, 10 hours post-extubation; Opioid consumption in first 24 hours (morphine equivalent); Time to Extubation (hours); ICU LOS; NRS Score Day 4 after Neuropathic Agent initiated; Breakthrough pain in first 48hrs (average #/pt);

SETTING: Medical and Surgical ICUs

PERSPECTIVE:


PRO

BLE

MIs the problem a priority?○ No○ Probably no○ Probably yes● Yes○ Varies○ Don't know

It is top priority to investigate the effectiveness/side effects of Adjunct Non opioid analgesics to opioids compared to opioids alone in order to decrease opioids related side effects that are especially at risk in ICU patients:- respiratory depression- neurological impairment (decreased vigilance, delirium)- ileus, nausea vomiting, etc...Also, opioids are not very effective to treat neuropathic pain compared to more specific analgesics (antineuropathic agents). If nociceptive pain is probably the most frequent mechanism of pain in ICU patients (medical, surgical illness and trauma; nociceptive care procedures), some patients can experiment a typical neuropathic pain as patients hospitalized in ICU for a Guillain-Barré syndrome.

Four single center double blinded RCTs:Patients with Guillain-Barré syndrome : 2 RCTsPandey 2002 [7](cross over design, n=18), Pandey 2005[8] (parallel groups, n=12 x 3)ICU patients admitted after elective cardiac surgery : 2 RCTsPesonen 1997 [9] (parallel groups, n=29 vs n=31)Joshi 2013 [10] (parallel groups, n=20 vs n=20)Drugs:Gabapentin vs. placebo (Pandey 2002) [7]Gabapentin vs. Carbamazepine vs. placebo (Pandey 2005) [8]Pregabalin vs. placebo (Pesonen 1997, Joshi 2013) [9, 10]OUTCOMES:1. Pain assessmentMean 0-10 NRS of the day till Day 7 (Pandey studies) [7, 8]Number of patients with VRS > 1 at rest (0-4 VRS) till 16-hrs after extubation; then at 1 and 3 months at rest and during movement = to assess chronic pain (Pesonen study)VAS (0-10), at rest and at deep breath till H48; then at 1 and 3 months at rest and during movement = to assess chronic pain (Joshi study)Forrest plot for NRS (0-10) = -3 [-4 to -3] (Pandey studies) [7, 8]NB: NRS is also significantly reduced in the gabapentin group compared to carbamazepine group in Pandey 2005.Pesonen study:Proportion of patient with VRS>1 at H16 postextubation: RR 0.3 (0.1 to 0.8)Proportion of patient with VRS>1 at 1 and 3 months: NS except at 3 months during movement:RR 0.14 (0.02 to 1.04) but crosses the line...Joshi study:VAS only provided as a figure, not computedBreakthrough pain in first 48hrs (average #/pt) : MD 1.45 lower (2.13 lower to 0.77 lower2. Opioid consumption in first 24 hours (morphine equivalent): 4 RCTs(Fentanyl in Pandey [7, 8] studies, oxycodone in Pesonen [9], tramadol in Joshi [10])MD 13.54 lower (14.57 lower to 12.5 lower)3. Time to extubation (hours): 2 RCTs in patients after elective cardiac surgery (Pesonen [9], Joshi [10])MD 0.36 higher (0.7 lower to 1.43 higher) -> NS

DES

IRA

BLE

EF

FEC

TSHow substantial are the desirable anticipated effects?○ Trivial○ Small○ Moderate● Large○ Varies○ Don't know

"large"because there is an important reduction in pain intensity as well as a reduction in opioid consumption- Moderate or Large

UN

DES

IRA

BLE

EF

FEC

TS

How substantial are the undesirable anticipated effects?○ Large○ Moderate● Small○ Trivial○ Varies○ Don't know

"trivial" - some undesirable effects not reported (e.g., dizziness)Bram: consider gap in the literature and those reported in selected studiesGroup agrees to "small"

CER

TAIN

TY O

F EV

IDEN

CE

What is the overall certainty of the evidence of effects?○ Very low○ Low● Moderate○ High○ No included studies

Pandey [7, 8] studies seem to have a low ROB but exclusion of patients after enrollment was not reported. Also, these two studies came from the same single center in India: Indian patients might have different pharmacoK/D than other populations? Any Pharm recommendation John?Pesonen study [9]: very high ROB: 10 on 70 patients were excluded, no intent to treat analysisNote that the population's size of each group in these 3 RCTs varies from n=12 to n=31 only.

"moderate" certainty of the evidence of effectsbecause 4 double blinded RCTs but with limitations-it should be whatever the harm of using the drug is - moderate is fair

VA

LUES


Neuropathic pain is difficult to treat with opioids. "No important" uncertaintyto be consistent with previous questions

BA

LAN

CE

OF

EFFE

CTS

Does the balance between desirable and undesirable effects favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison○ Probably favors the intervention● Favors the intervention○ Varies○ Don't know

"Favors" the interventionbecause there was a strong reduction in pain intensity, as well as a strong reduction in opioid consumption, a significant improvement in vigilance status and some improvement in CAM-ICU in one study (at Day 1 only).No other side effects when assessed, especially nausea, diarrhea, constipation.

RES

OU

RC

ES

REQ

UIR

ED

How large are the resource requirements (costs)?○ Large costs○ Moderate costs● Negligible costs and savings○ Moderate savings○ Large savings○ Varies○ Don't know

Costs of neuropathic agents are small in France. What in US and Canada?Low costs and accessible.

Negligible costs and savings

CER

TAIN

TY O

F EV

IDEN

CE

OF

REQ

UIR

ED R

ESO

UR

CESWhat is the certainty of the evidence

of resource requirements (costs)?○ Very low○ Low○ Moderate○ High● No included studies

CO

ST E

FFEC

TIV

ENES

S


No included studies(no cost-effectiveness study, to be consistent with other questions)

EQU

ITY


AC

CEP

TAB

ILIT

Y Is the intervention acceptable to key stakeholders?○ No○ Probably no○ Probably yes● Yes○ Varies○ Don't know

FEA

SIB

ILIT

Y


Neuropathic agents require the use of a gastric tube in patients unable to swallow, as well as the absence of ileus.

"varies"Yes or Probably yes when not focusing on specific ICU patients with CI Probably yes






studies

VALUESImportant


Possibly important





variability


BALANCE OF EFFECTS











COST EFFECTIVENESS







studies


Probably no impact




Should adjunct neuropathic agent vs. no adjunct neuropathic drug be used for ICU pain management (5.2)?TYPE OF RECOMMENDATION



intervention




intervention


○ ○ ○ ○ ●

RECOMMENDATION 1. We recommend to use adjunctive gabapentin to treat neuropathic pain and reduce opioid consumption in ICU patients with Guillain-Barré syndrome (+1B).2.1. We suggest that adjunctive neuropathic agents be used to treat neuropathic pain in ICU patients, i.e. patients with neuropathic pain related to ICU-aquired weakness (+2B).2.2. We suggest that adjunctive pregabalin might be used to decrease postoperative pain intensity and opioid consumption in some patients (+2B).have 2 recommendations: we recommend using neuropathic agents to treat neuropathic pain in the ICU (strong recommendation); weak recommendation for postoperative pain in CV surgery

JUSTIFICATION Four RCTs. Two in patients with Guillain-Barré syndrome, low ROB but from a single center and small size population; Two in patients after selective cardiac surgery, two centers, small size population, high ROB for one study.-> These RCTs are downgraded to a level of evidence of BAlso see recent metaanalyses:Pharmacological treatment for pain in Guillain-Barré syndrome. Jia Liu, Lu-Ning Wang, Ewan D McNicol. The Cochrane Library 2013 [11]Although reductions in pain severity were found when comparing gabapentin and carbamazepine with placebo, the evidence was limited and its quality very low. Larger, well-designed RCTs are required to further investigate the efficacy and safety of potential interventions for patients with pain in GBS. Additionally, interventions for pain in the convalescent phase of GBS should be investigated.Perioperative use of pregabalin for acute pain—a systematic review and meta-analysis. Naveen Eipea, John Penninga, Fatemeh Yazdib, Ranjeeta Mallickb, Lucy Turnerb, Nadera Ahmadzaib, Mohammed Toseef Ansarib. PAIN 2015 [12]Pregabalin analgesic effectiveness is largely restricted to surgical procedures associated with pronociceptive mechanisms. The clinical significance of observed pregabalin benefits must be weighed against the uncertainties about serious harms and enhanced recovery to inform the careful selection of surgical patients.


There is no data regarding pain in patients with ICUAW. However, there is an empirical evidence that these patients can suffer from typical neuropathic pain. Pathophysiological mechanism in ICUAW is likely similar to patients with Guillain-Barré syndrome. Further studies are needed to investigate pain in ICU patients with ICUAW, as well as the effect of adjunctive neuropathic agents to prevent and treat pain, and reduce opioid consumption.

RESEARCH PRIORITIES

This is top priority to investigate analgesics that can decrease the use or dose of opioids to treat pain in ICU patients as well as to investigate analgesics effective to treat neuropathic pain specifically.These results require that these studies be replicated:- in medical or surgical patients having neuropathic pain related to ICUAW- non communicant patients, i.e. at the early stage of sedation-analgesia


Can’t generalize GBS patient data to all critically ill patients.

Question: Adjunctive lidocaine compared to no lidocaine for ICU analgesic regime (5.2) Setting: Intensive Care Unit Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision


Adjunctive lidocaine

no lidocaine

Relative(95% CI)

Absolute

(95% CI)

VAS Score at 96 hours (follow up: mean 96 hours; assessed with: VAS; Scale from: 0 to 10)

1 randomised trials

serious a not serious not serious b serious c none 44 45 - MD 0.1 cm

higher(0.38

lower to 0.58

higher)

⨁⨁◯◯LOW

IMPORTANT

Total postoperative fentanyl required (assessed with: dosage (mcg))

1 randomised trials

serious a not serious not serious b serious c none 44 45 - MD 68.73 mcg

lower(387.372 lower to 249.912 higher)

⨁⨁◯◯LOW

IMPORTANT

CI: Confidence interval; MD: Mean differenceExplanationsa. unclear method of randomization, no intention to treat, exclusion of patients unclear b. Only included cardiac ICU patients; however effects may be generalizable c. Wide confidence intervals cannot distinguish if clinically important benefits with lidocaine adjunct

QuestionShould Adjunctive lidocaine vs. no lidocaine be used for ICU analgesic regime (5.2)?

POPULATION: ICU analgesic regime (5.2) BACKGROUND:

INTERVENTION:

Adjunctive lidocaine

COMPARISON: no lidocaine

MAIN OUTCOMES:

VAS Score at 96 hours; Total postoperative fentanyl required;


PERSPECTIVE:Assessment

JUDGEMENT RESEARCH EVIDENCE ADDITIONAL CONSIDERATIONS

PRO

BLE

M


It is top priority to investigate the effectiveness/side effects of Adjunct Non opioid analgesics to opioids compared to opioids alone in order to decrease opioids related side effects that are especially at risk in ICU patients:- respiratory depression- neurological impairment (decreased vigilance, delirium)- ileus, nausea vomiting, etc...One single center double-blinded RCT in 100 postcardiac surgery ICU patients, high ROB (No flow chart, exclusion of patients unclear, one patient in the intervention group was excluded from analysis because of death, no intent to treat analysis) -> Insler 1995 [13]NB: lidocaïne or placebo was begun just after induction of anesthesiaPain intensity at H96 (VAS in cm): MD 0.1 higher (0.4 lower to 0.6 higher)Fentanyl consumption (µg): MD 68 lower (39 lower to 249 higher)Ramsay score: Lidocaïne group significantly more sedated (higher score) at H4, not at H1, 2, 8, 16, 24, 48, 96.Other outcomes NS: hemodynamics, midazolam, propofol dosages, time to extubation, ICU LOS, hospital LOS

DES

IRA

BLE

EF

FEC

TS

How substantial are the desirable anticipated effects?● Trivial○ Small○ Moderate○ Large○ Varies○ Don't know

"trivial" because there was globally no (significant) effect at all.Also crossed the line of no effect.

UN

DES

IRA

BLE

EF

FEC

TSHow substantial are the undesirable anticipated effects?○ Large○ Moderate○ Small● Trivial○ Varies○ Don't know

"trivial"because there was globally no (significant) effect at all.

CER

TAIN

TY O

F EV

IDEN

CE

What is the overall certainty of the evidence of effects?○ Very low● Low○ Moderate○ High○ No included studies

Only one study, high ROB: ,o flow chart, exclusion of patients unclear, one patient in the intervention group was excluded from analysis because of death, no intent to treat analysis

"low"

VA

LUES


"Possibly important uncertainty or variability"because except fo hemodynamics (HR, MAP, PAP, CVP, CO, CI and monitoring for myocardial infarction by ECG and CPK enzymes), lidocaïne's related toxicity was not reported: arrhythmias, Central Nervous System (CNS) toxicity (seizures...)

- No uncertainty about VAS

BA

LAN

CE

OF

EFFE

CTS

Does the balance between desirable and undesirable effects favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison● Does not favor either the intervention or the comparison○ Probably favors the intervention○ Favors the intervention○ Varies○ Don't know

"Does not favor either the intervention or the comparison"

RES

OU

RC

ES

REQ

UIR

EDHow large are the resource requirements (costs)?○ Large costs○ Moderate costs● Negligible costs and savings○ Moderate savings○ Large savings○ Varies○ Don't know

Lidocaïne costs are small. "Negligible costs and savings"C

ERTA

INTY

OF

EVID

ENC

E O

F R

EQU

IRED

RES

OU

RC

ES

What is the certainty of the evidence of resource requirements (costs)?○ Very low○ Low○ Moderate● High○ No included studies

Old drug, not expensive and available-impact of nursing for cardiovascular assessment – recent meta-analyses available

CO

ST E

FFEC

TIV

ENES

S

Does the cost-effectiveness of the intervention favor the intervention or the comparison?○ Favors the comparison● Probably favors the comparison○ Does not favor either the intervention or the comparison○ Probably favors the intervention○ Favors the intervention○ Varies○ No included studies

No intervention "Favors the comparison"because adjunctive lidocaïne seems to be futile; not using IV lidocaïne saves the (small) cost of the drug but also the administration materials, as well as pharm and nurses' time.-Probably favors (only one study)

EQU

ITY

What would be the impact on health equity?○ Reduced○ Probably reduced○ Probably no impact○ Probably increased○ Increased○ Varies○ Don't know

AC

CEP

TAB

ILIT

Y Is the intervention acceptable to key stakeholders?○ No● Probably no○ Probably yes○ Yes○ Varies○ Don't know

"no"because adjunctive lidocaïne seems to be futile; not using IV lidocaïne saves the (small) cost of the drug but also the administration materials, as well as pharm and nurses' time.Probably no to be consistent with previous question

FEA

SIB

ILIT

Y


Take into account required monitoring






studies

VALUESImportant


Possibly important





variability


BALANCE OF EFFECTS











COST EFFECTIVENESS


Probably favors the

comparison





studies


Probably no impact




Should Adjunctive lidocaine vs. no lidocaine be used for ICU analgesic regime (5.2)?TYPE OF RECOMMENDATION


Conditional recommendation against

the intervention




intervention


○ ● ○ ○ ○

RECOMMENDATION We suggest against using adjunctive IV lidocaïne systematically in addition to opioids to treat pain and/or to reduce opioid consumption in postoperative ICU patients (-2C).Note: a weak recommendation does not imply that it cannot be used

JUSTIFICATION Only one low quality negative RCT in postcardiac ICU patients that investigated pain.-> level of evidence is downgraded to CAlso to take into consideration the recent Cochrane review (Continuous intravenous perioperative lidocaine infusion fo postoperative pain and recovery, by Kranke et al. Cochrane Library 2015)[14], Authors’ conclusions:There is low to moderate evidence that this intervention, when compared to placebo, has an impact on pain scores, especially in the early postoperative phase, and on postoperative nausea. There is limited evidence that this has further impact on other relevant clinical outcomes, such as gastrointestinal recovery, length of hospital stay, and opioid requirements. So far there is a scarcity of studies that have systematically assessed the incidence of adverse effects; the optimal dose; timing (including the duration of the administration); and the effects when compared with epidural anaesthesia.


Cardiac surgery patients



Lidocaïne's heart and Central Nervous System (CNS) toxicity needs to be investigate in critically ill patients.

RESEARCH PRIORITIES This is top priority to investigate analgesics that can decrease the use or dose of opioids to treat pain in ICU patients.There is only one negative study in postoperative ICU patients. A Cochrane review reported low to moderate evidence that IV lidocaïne reduced pain at the early stage of postoperative period, opioid consumption as well as improved bowel function, especially after abdominal surgeryFuture studies are required:- in SICU patients recovering from an abdominal surgery- in medical ICU, especially non communicant patients at the early stage of sedation-analgesia


specific contexts could be clinically useful: neuro-muscular and neuro patients (post-op or in general), recovering opiate addicts?"We do not suggest routine lidocaine" harmonizes with the Cochrane review’s modest benefit& this keeps the door open to considering its use. Surprising contrast between this and the ketamine recommendation (where there seemed to be as little data yet a recommendation)

Question: Adjunct NSAIDs compared to no adjunct NSAIDs for postoperative ICU patients (5.2) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision


adjunct NSAIDs

no adjunct NSAID

s

Relative

(95% CI)

Absolute(95% CI)

VAS Pain Score 24 hours postoperatively

2 randomised trials

serious a not serious not serious serious b none 104 53 - MD 0.35 lower


⨁⨁◯◯LOW

CRITICAL

Opioid Consumption (morphine equivalents)

2 randomised trials

serious a serious c not serious serious d none 49 53 - MD 1.61 lower

(2.42 lower to 0.8 lower)


CRITICAL

VAS Pain Score with Deep Breath at 6 hours

1 randomised trials

not serious

not serious not serious serious d none 21 22 - MD 1.3 lower(2.36 lower to

0.24 lower)

⨁⨁⨁◯MODERATE

CRITICAL

VAS Pain Score with Deep Breath at 24 hours

1 randomised trials

not serious

not serious not serious very serious d,e

none 21 22 - MD 0.6 lower(1.44 lower to 0.24 higher)

⨁⨁◯◯LOW

CRITICAL

CI: Confidence interval; MD: Mean differenceExplanationsa. No intention to treat in either analysis, unclear method of randomization b. Unclear if benefit with adjunctive NSAIDs based on meta-analysis c. Very high degree of inconsistency (Isquared = 92%) with non overlapping confidence intervals however both show benefit so lowered one level d. Small sample size e. Confidence Intervals include both harm and benefit

QuestionShould adjunct NSAIDs vs. no adjunct NSAIDs be used for postoperative ICU patients (5.2)?

POPULATION: postoperative ICU patients (5.2) BACKGROUND:

INTERVENTION:

adjunct NSAIDs

COMPARISON: no adjunct NSAIDs

MAIN OUTCOMES:

VAS Pain Score 24 hours postoperatively; Opioid Consumption (morphine equivalents); VAS Pain Score with Deep Breath at 6 hours; VAS Pain Score with Deep Breath at 24 hours;

SETTING:

PERSPECTIVE:


PRO

BLE

M


It is top priority to investigate the effectiveness/side effects of Adjunctive Non opioid analgesics to opioids compared to opioids alone in order to decrease opioids related side effects that are especially at risk in ICU patients:- respiratory depression- neurological impairment (decreased vigilance, delirium)- ileus, nausea vomiting, etc... Two single center RCT in ICU patients after surgerySmall size of every group in both studies: from 21 to 31 patientsHigh ROB for both studies: no intention to treat, unclear method of randomization- Hynninen 2000 : elective cardiac surgery, n=120 (4 groups: 3 different NSAIDs and placebo)- Oberhofer 2005 [15]: abdominal surgery, n=43 (2 groups: ketopropfen and placebo)Pain measurement:VAS 0-10 in Hyminnen at H3, 6, 12, 24 (at rest only)NRS 0-10 in Oberhoffer [15] at H3, 6, 12, 24 (at rest and deep breath)

Drugs:Hynninen:Adjunctive 75 mg diclofenac or 100 mg ketoprofen or 100 mg indomethacin or placebo (4 groups) one hour before intubation (note: how is it possible to know the extubation time exactly!?!) and 12 hours afterwardBasal analgesia: none?Rescue analgesia: morphine in the ICU (and mepiridine for shivering); paracetamol, codeine in the ward (transfer on Day 1)Oberhofer:Adjunctive 100 mg ketoprofen IV one and nine hours postopBasal analgesia: tramadol and metamizoleRescue analgesia: tramadol 25 mg IV

DES

IRA

BLE

EFF

ECTS



Forest plot for pain scales (0-10) at rest: MD 0.35 fewer (0.91fewer to 0.21 more)NB: data are provided only in a figure of means without SD for Hynninen, not estimable for the forrest plot ; in the body text: non significant difference at all times.Pain at deep breath (Oberhofer) [15]:H6: MD 1.3 fewer (2.36 fewer to 0.24 fewer)H 24: MD 0.6 fewer (1.44 fewer to 0.24 more)

Forest plot for morphine consumption at H24 (mg equivallent):MD 1.61 fewer (2.42 fewer to 0.8 fewer)Other outcomes:- Nausea and/or vomiting: non significant difference in NSAIDs groups versus placebo (both studies) but very small size groups, probable lack of power- Other opioid related effects not reported, including: sedation, ileus, duration of MV, LOS in ICU... only "respiratory depression" was reported in Oberhofer [15]: no case in any group- Serum creatinine: non significant change between groups but one patient had an important increase in S-creatinine after the first dose of indomethacine and was withdrawn of Hynninen's study (no intent to treat analysis!)- No significant difference in excessive bleeding but again, a probable lack of power, also note that one patient in the indomethacin group had excessive bleeding and was withdrawn of Hynninen's study (no intent to treat analysis!)- No platelet anti-agregant test was performed to show any effect on platelet function- Incidence of peptic ulcer in the postoperative period was not reported

"small" desirable anticipated effectsbecause non significant reduction in pain intensity at rest H24, a significant reduction at deep breath in one study, a very small reduction in opioid consumption at H24

UN

DES

IRA

BLE

EFF

ECTS

How substantial are the undesirable anticipated effects?○ Large○ Moderate○ Small○ Trivial

○ Varies● Don't know

don't know" undesirable anticipated effectsbecause non significant difference for adverse outcomes but two severe adverse events possibly related to NSAID among 30 patients who received indomethacin, no intent to treat analysis; no platelet functional test used on bothe studies

CER

TAIN

TY O

F EV

IDEN

CE

What is the overall certainty of the evidence of effects?○ Very low● Low○ Moderate○ High○ No included studies

Small size of every group in both studies: from 21 to 31 patientsHigh ROB for both studies: no intention to treat, unclear method of randomization- Hynninen: 6 patients were excluded among 120, included two for indometacine possible side effect- Oberhofer [15]: no patient seems to be excluded after enrollement in the intervention group, one for rash in the control group

"low" certainty of the evidence of the effects

VA

LUES


"no important" uncertainty about or variability in how much people value the main outcomes to be consistent with previous questions

BA

LAN

CE

OF

EFFE

CTS

Does the balance between desirable and undesirable effects favor the intervention or the comparison?○ Favors the comparison○ Probably favors the comparison○ Does not favor either the intervention or the comparison○ Probably favors the intervention○ Favors the intervention○ Varies● Don't know

"don't know" if the balance between desirable and undesirable effects favor the intervention or the comparisonbecause of no or small benefits (no reduction in pain intensity at rest at H24, very small reduction in opioid consumption within 24h) and don't know harms

RES

OU

RC

ES

REQ

UIR

ED

How large are the resource requirements (costs)?○ Large costs○ Moderate costs○ Negligible costs and savings○ Moderate savings○ Large savings○ Varies● Don't know

NSAIDs costs are small. "don't know" about the resource requirements because severe adverse events related to drugs can be very expensive. See assessment of undesirable effetcs.

CER

TAIN

TY O

F EV

IDEN

CE

OF

REQ

UIR

ED R

ESO

UR

CES

What is the certainty of the evidence of resource requirements (costs)?○ Very low● Low○ Moderate○ High○ No included studies

"low" certainty of the evidence of ressource requirementsSee assessment of undesirable effetcs.

CO

ST E

FFEC

TIV

ENES

S


No included studies because no cost-effectiveness study, potential unassessed harms.

EQU

ITY


AC

CEP

TAB

ILIT

Y Is the intervention acceptable to key stakeholders?○ No● Probably no○ Probably yes○ Yes○ Varies○ Don't know

intervention acceptable to stakeholders -> "probably NO"because no or small benefits and potential harms related to the intervention.

FEA

SIB

ILIT

YIs the intervention feasible to implement?○ No○ Probably no● Probably yes○ Yes○ Varies○ Don't know






studies

VALUESImportant


Possibly important





variability


BALANCE OF EFFECTS











COST EFFECTIVENESS







studies


Probably no impact




ConclusionsShould adjunct NSAIDs vs. no adjunct NSAIDs be used for postoperative ICU patients (5.2)?TYPE OF RECOMMENDATION Strong

recommendation against the intervention

Conditional recommendation

against the intervention


either the intervention or the comparison


the intervention

Strong recommendation for

the intervention

○ ● ○ ○ ○

RECOMMENDATION We suggest against using adjunctive NSAIDs to prevent pain in ICU patients (-2C).

JUSTIFICATION Overall justificationTwo RCTs but single centers, high ROB, small size-> level of evidence downgraded to CDetailed justificationProblem Surgical ICU population (one cardiac, one abdominal). However, generalization of the results to another kind of population is likely because pain is frequent in SICU patients as in medical ICU patients.Desirable Effects No or small desirable effectsUndesirable Effects Some severe adverse events possibly related to NSAID reported in one study, no intent to treat analysis, small size studies that cannot conclude about these adverse events



MONITORING AND EVALUATION Analgesic related side effects or outcomes should be assessed more largely including:-delirium-ileus-duration of MV weaning-LOS in ICU and hospital NSAIDs related side effects should be assessed more precisely:- postoperative bleeding should be defined according to standardized definitions- risk of bleeding should be assessed by anti-agregant platelet tests, instead of activated coagulation time- peptic ulcer

RESEARCH PRIORITIES This is top priority to investigate analgesics that can decrease the use or dose of opioids to treat pain in ICU patients.This kind of study needs to be replicated: - in medical ICU- non verbal patients


Single RCT; 1-2 doses likely safe/ effective, more may shift the benefit/burden ratioadd "non-procedural pain" as we for procedural pain.

Question: Cybertherapy compared to no cybertherapy for ICU pain management (5.3) Setting: Non-pharmacological analgesic therapies Bibliography:

Quality assessment

Impact Quality Importance

№ of studie

s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other consideration

s

# of patients with reduced pain post-cybertherapy

1 observational studies

not serious

not serious not serious serious a none Mosso-Vazques [16] used virtual reality cybertherapy to reduce postoperative pain in cardiac surgical ICU patients. 67 patients were monitored within the first day postoperatively. Pain scores were obtained pre- and post-cybertherapy. 59/67 patients (88%) reported a decreased level of pain, with a mean Likert scale change of 3.75, corresponding with a change from severe to moderate, or moderate to light. Physiologically, 25 (37.3%) patients experienced reduced heart rates, 35 (52.2%) experienced reduced MAPs, and 14/22 (64% tested for respiratory rates experienced a reduction.


CRITICAL

CI: Confidence intervalExplanationsa. Small sample size

Question: Should cybertherapy vs. no cybertherapy be used for ICU pain management (5.3)?

Population: ICU pain management Background:

Intervention: cybertherapy

Comparison: no cybertherapy

Main outcomes: # of patients with reduced pain post-cybertherapy

Setting: Non-pharmacological analgesic therapies

Perspective:

Assessment

Criteria Judgements Research evidence Additional considerations

Problem Is there a problem priority?

○ No ○ Probably no ○ Uncertain ○ Probably yes ● Yes ○ Varies

Non-pharmacological interventions for pain management are important in the ICU and could contribute to reduce pain and to improve pain control in patients.

Benefits & harms of the options

What is the overall certainty of this evidence?

○ No included studies ● Very low ○ Low ○ Moderate ○ High

Summary of evidence:Only 1 observational study (pre/post) was included (Mosso-Vazquez et al., 2014) [16]. The aim was to evaluate the use of virtual reality (VR) cybertherapy on postoperative distress in cardiac surgery ICU patients. 67 patients navigated a 30-minute VR simulation designed for pain management. 59

The relative importance or values of the main outcomes of interest:

Outcome Relative importance

Certainty of the evidence

(GRADE)

# of patients with reduced pain post-cybertherapy

CRITICAL ⨁◯◯◯VERY LOW

patients (88%) reported a decreased level of pain post-therapy. The mean change was 3.75 on a Likert scale (unsure what the possible range was). The authors described that change as a decrease from severe to moderate or moderate to liht. Methods section very brief and poorly written.Uncertainty about main outcomes:-unsure what the Likert scale scores are.-important uncertainty or variability-important uncertainty or possibly importantAre the desirable anticipated effects large?I put uncertain because we only have 1 observational study, the sample size was small and the method is questionable. On the other hand, a change in 3.75 would be considered clinically significant if we would know what the scores mean (Likert scale).Group: uncertain - we don't really knowAre the undesirable anticipated effects small?Complications (6% of patients) during treatment included cardiac arrhythmia (n=1), and nausea and vertigo (n=3) that interrupted cybertherapy session (p.375). Based on this information, I am hesitant between Probably no and Uncertain.- these complications could happen anytime-happen frequently in cardiac ICU patients-uncertain because could happen during the intervention too; number is too small to make a judgmentConsensus from the group for uncertainAre the desirable effects large relative to undesirable effects?Uncertain because unsure about the desirable effects and complications were reported (but in a small number of patients).Consensus from the group for uncertain

Is there important uncertainty about how much people value the main outcomes?

● Important uncertainty or variability ○ Possibly important uncertainty or variability ○ Probably no important uncertainty or variability ○ No important uncertainty or variability ○ No known undesirable outcomes

Are the desirable anticipated effects large?

○ No ○ Probably no ● Uncertain ○ Probably yes ○ Yes ○ Varies

Are the undesirable anticipated effects small?


Are the desirable effects large relative to undesirable effects?


Resource use Are the resources required small?

○ No ● Probably no ○ Uncertain ○ Probably yes ○ Yes ○ Varies

A head-mounted display was installed on the patient's head to display a VR simulation. A projector emitted the same simulation on the unit wall. The simulation consisted of five cybertherapy environments. The simulation lasted 30 minutes.

Special equipment is necessary for this intervention and time and resources to develop the simulations should also be taken into account.- addresses music therapy as an example and all the necessary resources needed- No - enormous amount of resources`probably no also acceptable-Probably no- Probably noPt: Probably no

Is the incremental cost small relative to the net benefits?


RCTs are necessary to confirm conclusions on the effectiveness of this intervention on pain management.Costs for this intervention should be considered.We cannot conclude that benefits exceeded the costs.Consensus from the group for uncertain.

Equity What would be the impact on health inequities?

○ Increased ○ Probably increased ○ Uncertain ○ Probably reduced ○ Reduced ○ Varies

Acceptability Is the option acceptable to key stakeholders?


The ICU environment and access to equipment and resources will likely influence the acceptability of this intervention by stakeholders.

Consensus from the group for uncertain

Feasibility Is the option feasible to implement?


Many factors (equipment, time, resources, ICU environment, training) should be taken into account.Consensus for probably no

RecommendationShould cybertherapy vs. no cybertherapy be used for ICU pain management (5.3)?

Balance of consequences

Undesirable consequences clearly outweigh desirable

consequences in most settings

Undesirable consequences probably outweigh desirable


The balance between desirable and undesirable consequences

is closely balanced or uncertain

Desirable consequences probably outweigh

undesirable consequences in most settings

Desirable consequences clearly outweigh undesirable


○ ○ ● ○ ○

Type of recommendation We recommend against offering this option

We suggest not offering this option We suggest offering this option We recommend offering this option

○ ● ○ ○

Recommendation We recommend against offering VR cybertherapy until further research is available.From Bram: I agree with your assessments. You may consider recommending more research and in the meantime recommending against (conditionally) until more research is available. Alternatively you could just recommend further work without making a formal recommendation. My slight preference would be for the former.

Justification I believe because of the very low evidence and the complications reported, the balance of desirable and undesirable consequences is uncertain. But maybe the reduction in pain level probably outweigh the risks. I would be in favor of suggesting to conduct RCTs to allow a higher level of evidence to conclude on the effectiveness of such an intervention.

Subgroup considerations Only 1 observational study conducted with cardiac surgery ICU patients. Small sample size. No inferential statistics.

Implementation considerations

Many factors to be taken into account including the ICU environment, access to technology, resources. and training

Monitoring and evaluation

We need RCTs to draw firm conclusions about the effectiveness of such an intervention.

Research possibilities Rationale for recommendation: There is a major gap in research (1 observational study). Measure issue for pain (Likert scale). Small sample size (no power analysis) and inferential statistics not used.Other virtual interventions could be developed and tested. The use of touchpad or tablet could be considered.

Question: Hypnosis + pharmacological therapies compared to no hypnosis for critically ill pain management patients (5.4) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other consideration

s

Hypnosis + pharmacologica

l therapies

no hypnosi

s

Relative(95% CI)

Absolute

(95% CI)

Daily VAS Score (cm)


very serious a

not serious not serious very serious b,c

none 23 23 - MD 0.5 lower(1.37

lower to 0.37

higher)


CRITICAL

Morphine Equivalents


very serious a

not serious not serious very serious b

none Pre-post morphine equivalent dosing was not provided for the control arm. Intervention arm did have a large decrease in morphine equivalents between days 10-15; the authors attribute this to the completion of surgeries and wound management in these burn patients.


CRITICAL

Propofol Requirements (mg)


very serious a

not serious not serious very serious b

none 23 23 - MD 380 mg

lower(523

lower to 237

lower)


IMPORTANT

CI: Confidence interval; MD: Mean differenceExplanationsa. Inaccurate assessment of outcomes, variability on co-interventions between groups, unclear ascertainment of exposure b. Small sample size (only 23 patients eligible) c. CI cross line of no effect

Question Should Hypnosis + pharmacological therapies vs. no hypnosis be used for critically ill pain management patients (5.4)?

Population: critically ill pain management patients (5.4)

Background:

Intervention: Hypnosis + pharmacological therapies

Comparison: no hypnosis

Main outcomes: Daily VAS Score (cm) Morphine Equivalents Propofol Requirements (mg)

Assessment




Non-pharmacological interventions are important to consider in ICU pain management.



○ No included studies ● Very low ○ Low ○ Moderate ○ High

Only one small observational study (with matched controls, n=23 in each group); many limitations.Are the desirable anticipated effects large?



Certainty of the evidence

(GRADE)

Daily VAS Score (cm)


Morphine Equivalents


Propofol Requirements (mg)

IMPORTANT ⨁◯◯◯VERY LOW

Pain score MD very small (0.5 lower) and CI cross line of no effect.Are the undesirable anticipated effects small?Uncertain; adverse events not reported. No difference in vital signs (HR, BP, RR), and first bowel movement between groups.Are the desirable effects large relative to undesirable effects?Probably no, but could put uncertain. We cannot conclude to small desirable effect considering that small observational study, and undesirable effects remain unknown (but unlikely).


○ Important uncertainty or variability ○ Possibly important uncertainty or variability ○ Probably no important uncertainty or variability ● No important uncertainty or variability ○ No known undesirable outcomes









In this study, hypnosis was administered by ICU nurses who had completed 3 years of training under the supervision of a psychiatrist.



Training appears extensive.Intervention with many stages; duration may vary between patients.


○ Increased ○ Probably increased ● Uncertain ○ Probably reduced ○ Reduced ○ Varies

Unlikely



Uncertain based on required resources.



Feasibility concernsWill vary between patients (hypnosibility)

RecommendationShould Hypnosis + pharmacological therapies vs. no hypnosis be used for critically ill pain management patients (5.4)?












○ ○ ● ○ ○


We suggest not offering this option

We suggest offering this option

We recommend offering this option

○ ● ○ ○

Recommendation No recommendation? The group feels more confortable to not make a recommendation and document why we cannot make recommendation.

Methods- I would steer the group towards a conditional/weak recommendation against given the lack of benefit (acknowledging all very low quality evidence) and the costs/resources involved in training.I believe there is utility in making a recommendation versus not making one. And I think here a weak recommendation against makes sense - if the group agrees. It doesn't mean we're discouraging future research work and we can be clear about that (always everyone's worry in this case).

Justification Little evidence to support a recommendation for hypnosis.In the justification you can obviously elaborate and say that if future evidence shows benefit this balance may change or in select centers with expertise the balance may be different.

Subgroup considerations The included study was conducted in a burn ICU during dressing and hydrotherapy.


Not yet to be considered for implementation, need stronger evidence and RCTs.


Research possibilities Ideas form the group?Having higher level of evidence, include other ICU patients and during care procedures. Review research in anesthesia (morphine sparing effects, anesthesia) and specific surgical procedures (pediatric and gynecological).

Comments during electronic voting by entire panel

Hard to imagine virtual reality as good option for a population at such high risk for delirium (or certainly an altered state of consciousness).

Question: A narcotic compared to placebo for critically ill adults undergoing a procedure (2.1) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other considerations a narcotic placebo Relative

(95% CI)

Absolute

(95% CI)

Pain Score

2 randomised trials

not serious

not serious a not serious serious b none 59 56 - SMD 1.27 SD lower(2.79

lower to 0.26

higher)


CRITICAL

CI: Confidence interval; SMD: Standardised mean differenceExplanationsa. High Isquared (>90%), although both studies show benefit. b. Confidence intervals don't exclude harm.

Question Should a narcotic vs. placebo be used for critically ill adults undergoing a procedure (2.1)?

Population: critically ill adults undergoing a procedure (2.1)

Background:

Intervention: a narcotic

Comparison: placebo

Main outcomes: Pain Score

Assessment






○ No included studies ○ Very low ○ Low ● Moderate ○ High



Certainty of the evidence (GRADE)

Pain Score

CRITICAL ⨁⨁⨁◯MODERATE

Two studies included.Robleda et al (2015)[17] : Randomized, double-blind, paralell group, placebo-controlled clinical trial, 2 groups: 1) fentanyl (1 ug/kg for medical patients and 1.5 ug/kg for surgical/trauma patients) (n=39); and 2) control (n=36).Procedure: turning in mechanically ventilated patients (heterogeneous group)Pain measure: BPS ( a score >3 reflects pain, and a score >5 indicates significant pain)

Main findings: Pain incidence rate was significantly lower in the fentanyl group (74%) compared to control group (94%) with a relative risk of 0.79; incidence of significant pain was not statistically different between groups (49% versus 64%) with a relative risk of 0.76. Magnitude of pain (AUC): the fentanyl group had a lower magnitude of pain than the control group during turning and 30 minutes after. Adverse events and vital signs also reported.Casey et al. (2010): Randomized, double-blind clinical trial, 3 groups: 1) remifentanyl 1 ug/kg (n=20), 2) remifentanyl 0.5 ug/kg (n=20), and 3) placebo (n=20). I assume the remifentanyl 0.5 ug/kg was only included in the evidence table (to be confirmed with John C and Bram).Procedure: CTR in cardiac surgery ICU patientsPain measure: 10 cm VASMain findings: both remifentanyl groups had no increase in pain scores during CTR compared with baseline (median=1 for remifentanyl 5 ug/kg and 0 for remifentanyl 1 ug/kg; the median was 2 for both groups at baseline); the placebo group showed an increase in VAS pain scores during CTR (median=5). Vital signs data (SBP, HR, RR, SpO2) also reported.Overall SMD was 1.27 lowerAre the desirable anticipated effects large?Probably yes because did not reach clinical significance




○ No ○ Probably no ○ Uncertain ● Probably yes ○ Yes ○ Varies


○ No ○ Probably no ○ Uncertain ○ Probably yes ○ Yes ● Varies

reduction in pain (2 cm). More difficult to conclude with BPS, but pain incidence (BPS>3) was lower in the fentanyl group by 20%.Paul, Kathleen and JF: BPS of 3 indicates no pain behaviors were present? Not necessarily... Nurses can administer analgesics if BPS >4 (JFAre the undesirable anticipated effects small?Probably yes because no serious adverse events (but 4 patients with respiratory depression in the fentanyl group) reported in Robleda study; 19 non-serious events (e.g., transient hypotension, vomiting) with no significant differences between groups. In Casey study, reductions in all vital signs in the remifentanyl 1 ug/kg group, but only for blood pressure and RR in the remifentanyl 0.5 ug/kg group. No differences in vital signs





Opioids are part of standard of care. However, remifentanyl is more expensive compared to fentanyl.



Probably yes because opioids are easily accessible in the ICU, and are shown to be effective in reducing pain scores during procedures.Remifentanil more expensive compared to fentanyl.


○ Increased ○ Probably increased

○ Uncertain ○ Probably reduced ○ Reduced ○ Varies



Probably yes for fentanyl. For remifentanil, it may vary in different countries based on costs and standard practice.



Probably yes for fentanyl, but uncertain for remifentanil based on my comments above.-: Probably yes considering the 2 trials together- feasible and quick to actionPt-: adverse effects lasting shorter

RecommendationShould a narcotic vs. placebo be used for critically ill adults undergoing a procedure (2.1)?












○ ○ ○ ● ○





○ ○ ● ○

Recommendation We suggest using an opioid prior to a procedure (e.g., CTR, turning) for pain management in ICU patients (+2B).

Justification Well designed RCTs showing reductions in pain scores or pain incidence when opioids are administered compared to a placebo.

Subgroup considerations CTR in cardiac surgery ICU patients (Casey) [18]; turning in an heterogeneous ICU patient group (Robleda)[17]


Costs and standard practice in different countries; timing of administration prior to the procedure is important to consider (but quick to act); adverse events may occur and should be monitored and treated in a timely manner


Pain assessment prior and during (or right after the procedure) is key for the evaluation of the effectiveness of analgesia; monitoring of adverse events should also be done

Research possibilities Other procedures and other opioids should be tested; for turning, it would be relevant to also include the self-report of pain intensity as a primary outcomeImplementation studies are also needed

Question: High dose narcotic compared to low dose narcotic for critically ill adults undergoing a procedure (2.1) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision


high dose narcotic

low dose narcotic

Relative(95% CI)

Absolute

(95% CI)

Pain Score (assessed with: SMD)

2 randomised trials

not serious

serious a not serious serious b none 78 79 - SMD 0.26 SD lower(0.94

lower to 0.42

higher)

⨁⨁◯◯LOW

CRITICAL

CI: Confidence interval; SMD: Standardised mean differenceExplanationsa. High Isquared (70%) and non-overlapping confidence intervals. b. Wide confidence intervals don't exclude harm.

Question

Should high dose narcotic vs. low dose narcotic be used for critically ill adults undergoing a procedure (2.1)?


Background:

Intervention: high dose narcotic

Comparison: low dose narcotic


Assessment




Pharmacological interventions for procedural pain management are important given the documented prevalence of pain associated with many ICU procedures.What is the overall certainty of this evidence? Casey 2010 [18]: significantly less pain with CTR or turning after remifentanyl 1 mcg/kg remifentanil IV (n=20) vs remifentanyl 0.5 mcg/kg remifentanil IV (n=20).CTR pain intensity scores per 0-10 VAS:Remifentanil 1 mcg/kg = median 0 [0-2]Remifentanil 0.5 mcg/kg = median 1 [0, 2]SMD -0.66 [-1.30, -0.02]Ahlers 2012 [19]: found no difference in pain with CTR with morphine 7.5 mg IV 30 minutes prior to procedure (n-58) vs morphine 2.5 mg IV 30 minutes prior to procedure (n=59). (Ahlers did not differentiate between those who were turned vs. turned and CTR.)Procedural pain scores per 0-10 NRS:Morphine 7.5 mg IV = mean 2.7 + 2.0-3.4Morphine 2.5 mg IV = mean 2.6 + 2.0-3.2SMD 0.04 [-0.32, 0.40]Overall SMD -0.26 [-0.94, 0.42] NSSo, only 1 small-sample study with significant findings.



○ No included studies ○ Very low ● Low ○ Moderate ○ High


Outcome

Relative

importance

Certainty of

the eviden

ce (GRADE)

Pain Score

CRITICAL

⨁⨁◯◯LOW

Are the desirable anticipated effects large? NoAre the undesirable anticipated effects small? Casey [18] had 2/20 patients with 1-3 minutes of apnea after 1 mcg/kg remifentanil IV, requiring bag and mask ventilation 1 minute after drug administration.Are the desirable effects large relative to undesirable effects? Depends on drug dosage, as in Casey [18].-Probably no considering that desirable effects are absent.- No (no positive effects)




● No


○ No ○ Probably no ● Uncertain


● No ○ Probably no ○ Uncertain ○ Probably yes ○ Yes ○ Varies

Resource use

Are the resources required small?


Administration of opioids is standard practice in ICUs.



Pain relief from opioid administration is low cost.=: Uncertain; Remifentanyl more expensiveGroup consensus



Acceptability

Is the option acceptable to key stakeholders?


Administration of opioids is standard practice in ICUs.- Uncertain because administration of opioids prior to a procedure raise many concerns among ICU clinicians.- Low versus high doses; so probably no (with high doses).- Probably no

Feasibility

Is the option feasible to implement?


Administration of opioids is standard practice in ICUs. However, administration of opioids prior to procedures requires ICU clinicians to probably change practice and plan for this.- Uncertain with only 2 studies and small samples in a homogeneous population.-Uncertain; will depend on intubation status and communication between nurses and physicians

Recommendation Should high dose narcotic vs. low dose narcotic be used for critically ill adults undergoing a procedure (2.1)?






The balance between desirable and undesirable consequences is closely

balanced or uncertain

Desirable consequences probably outweigh undesirable




○ ○ ● ○ ○

Type of recommendation

We recommend against offering this option We suggest not offering this option



○ ○ ● ○

Recommendation A substantial body of research demonstrates the painfulness of turning and chest tube removal. Pain should be assessed prior to a procedure, and administration of analgesics to address baseline as well as procedural pain should be made a standard part of practice.Opioids in low doses should be considered for procedural pain management. (in both studies, the scores were low no matter low or high doses were used)We suggest using opioids for chest tube removal. If using opioids, we suggest using low doses versus high doses. Little benefit and use of low versus high doses of opioids; I would suggest not offering this option. Recommending lower doses would be a better option. No benefit to administer opioids when pain at baseline is low. No control groups (with no opioids) in these studies For semantics, better to use positive than negative.

Justification 2013 PAD Guidelines [20] recommended use of analgesics prior to CTR.

Subgroup considerations

Ahlers [19] did not differentiate between patients being turned and patients being turned and having CTR. This confounds study findings. However, Ahlers said that patients just having CTR had significantly less pain than patients just being turned.


Administration of opioids prior to a procedure should be timed to the analgesic's peak effect, occurring at time of procedure.


Pre and post pain assessments to evaluate impact of the intervention. Potential adverse effects from opioids should be monitored.

Research possibilities

Study one procedure e.g., CTR using equianalgesic dose of remifentanyl 0.5 mcg/kg IV in a large sample. Alternative opioids.


RCT showing fentanyl (1-1.5 mcg/kg/ dose) before turning in mixed, MV population ↓ pain by BPS by 20%, NNT 5 (Robleda ICM 2016); 10% got respiratory depression.Caveats: Forrest plots (FPs) mitigated by outcomes combined from different studies; #1 improved pain scores opioid vs. placebo, #2 pain score differences with opioids + turning= no difference, also a positive outcome. But combining these different outcomes misleads the FPs. In which pt. population is turning painful?

Question: Local analgesia compared to nitrous oxide for critically ill adults undergoing a procedure (2.1) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision


local analgesia

nitrous oxide

Relative(95% CI)

Absolute

(95% CI)

Pain Score

1 randomised trials

serious a not serious not serious serious b none 22 22 - MD 27.5 lower(40.9

lower to 14.1

lower)

⨁⨁◯◯LOW

CRITICAL

New outcome

not estimable

-

CI: Confidence interval; MD: Mean differenceExplanationsa. No clear description of allocation concealment or blinding in single included study. b. Confidence intervals don't include no effect however small number of patients.

Question

Should local analgesia vs. nitrous oxide be used for critically ill adults undergoing a procedure (2.1)?


Background:

Intervention: local analgesia

Comparison: nitrous oxide


Assessment




Pharmacological interventions for procedural pain management are important given the documented prevalence of pain associated with many ICU procedures.




○ Very low ● Low ○ Moderate ○ High

Are desirable anticipated effects large?Only one study: Akrofi, 2005 [21]. CTR after cardiac surgery. Either 20 ml of 0.5% bupivacaine SQ infiltration around drain insertion site (comparator) (n=22) or inhaled 50% nitrous oxide and oxygen (Entonox) (control) (n=22). Median pain intensity score 9.5 (3–18) in bupivacaine group vs 37.0 (13–56) in




Pain Score

CRITICAL ⨁⨁◯◯LOW

Entonox group (p<0.05). SMD; - 27.50 [-40.90, -14.10] on 0 - 100 VAS. Large clinical significance. No clear description of allocation concealment or blinding in single included study.

Are the undesirable anticipated effects small? No difference among groups in arterial blood pressure, heart rate, PaCo2, oxygenation, or sedation.

- : Adverse events less likely to occur with local administration and low doses. Probably yes


○ Important uncertainty or variability ○ Possibly important uncertainty or variability ○ Probably no important uncertainty or variability ● No important uncertainty or variability




○ No ○ Probably no ○ Uncertain ● Probably yes





Re. Entonox use: While they are a part of intra-operative anesthesia, use of gas in ICU for procedural pain relief is unusual. And it was not effective. Re. bupivacaine use: not a large effort needed to administer subcutaneously around chest tube.



Could be part of standard practice





A qualified clinician would have to do the subcutaneous infiltration of bupivacaine.



Medication already available in clinical practice.

RecommendationShould local analgesia vs. nitrous oxide be used for critically ill adults undergoing a procedure (2.1)?












○ ○ ○ ● ○





○ ○ ● ○

Recommendation Bupivacaine infiltration is preferred over nitrous oxide prior to CTR.

Justification Effective in reducing CTR pain. Clinically significant pain decrease.

Subgroup considerations This study included removal of mediastinal chest tubes in cardiac surgery ICU patients. The effect on pleural tubes, whose removal is known to be more painful, is unknown.


Requires clinician qualified to do infiltration.

Monitoring and evaluation Pre- and post- CTR pain scores should be ascertained.

Research possibilities Replication of this study with a larger sample with heterogeneous patients and inclusion of a pleural chest tube group. This study had no clear description of allocation concealment or blinding in single included study. A future study should address these issues. Larger sample with enough power to determine differences in adverse effects.


should an opioid be used instead? Only comparing to opioids?

Question: A narcotic compared to nitrous oxide for critically ill adults undergoing a procedure (2.1) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other considerations a narcotic nitrous

oxideRelative

(95% CI)

Absolute

(95% CI)

Pain Score

1 randomised trials

serious a not serious not serious serious b none 22 22 - MD 22 lower(36.68

lower to 7.32

lower)

⨁⨁◯◯LOW

CRITICAL

CI: Confidence interval; MD: Mean differenceExplanationsa. No clear description of allocation concealment or blinding in single included study. b. Small number of patients - despite confidence intervals all on side of benefit.

Question

Should a narcotic vs. nitrous oxide be used for critically ill adults undergoing a procedure (2.1)?


Background:


Comparison: nitrous oxide


Assessment




Pharmacological interventions for procedural pain management are important given the documented prevalence of pain associated with many ICU procedures.




○ Very low ● Low ○ Moderate




Pain Score


Only one study: Akrofi, 2005 [21]. CTR after cardiac surgery. Either 50% nitrous oxide and oxygen (Entonox) (control) 2 minutes before procedure (n=22) or 0.1 mg/kg morphine as an IV bolus over 2 min 20 minutes before procedure (intervention)(n=22.)Median VAS scores on 0 - 100 scale: nitrous oxide group: 37.0 [13–56] vs. morphine group: 15.0 [7–27]SMD: -22 [-36.68, -7.32] Acceptable clinical significance. No clear description of allocation concealment or blinding in single included study.Are anticipated effects large? Uncertain or probably no given one study, pain scores not in the moderate-severe levels.Do we want one of these interventions to have a big difference on pain scores? Nitrous oxide patients had twice more pain than the morphine patients.Pamela: No differences with 2 and 4/10; side effects not considered.Gerald: One small study; 37 is pretty high so anticipated effects appear to be large. Could patients with nitrous oxide reliably self-report pain?-Not considering the sample size and side effects, just looking at this specific question what is the benefit of opioids large? With group: Considering this, probably yes makes sense.Are the undesirable anticipated effects small? No difference among groups in arterial blood pressure, heart rate, PaCo2, oxygenation, or sedation.Desirable versus undesirable effects: Changed to probably yes


● No important uncertainty or variability


● Probably yes ○ Yes ○ Varies


● Yes ○ Varies





Use of IV opioids is part of standard practice.



No incremental costs.


○ Increased ○ Probably increased ○ Uncertain ○ Probably reduced ○ Reduced X○ Varies



Use of IV opioids is part of standard practice.



RecommendationShould a narcotic vs. nitrous oxide be used for critically ill adults undergoing a procedure (2.1)?












○ ○ ○ ● ○


We recommend against offering this option




○ ○ ● ○

Recommendation We suggest to premedicate patients with opioids instead of nitrous oxide prior to mediastinal chest tube removal.

Justification Clinical significance difference in favor of morphine; small study

Subgroup considerations

This study included removal of mediastinal chest tubes in cardiac surgery patients. The effect on pleural tubes, whose removal is known to be more painful, is unknown.Use of short-term opioids (e.g., fentanyl, alfentanyl, remifentanyl) to manage short-duration procedures.


Opioids are already implemented in standard practice.


Pain assessment prior to the procedure with validated scales. Side effect monitoring.

Research possibilities Replication of this study with a larger sample and inclusion of a pleural chest tube group. This study had no clear description of allocation concealment or blinding in single included study. A future study should address these issues.Larger sample size necessary to capture adverse events. As a patient, I would prefer nitrous oxide. Differences among different types of opioids in terms of side effects.

Question: Isoflurane & Nitrous Oxide compared to Nitrous oxide alone for critically ill adults undergoing a procedure (2.1) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision


Isoflurane & Nitrous

Oxide

Nitrous oxide alone

Relative

(95% CI)

Absolute

(95% CI)

Pain Score

1 randomised trials

serious a not serious serious b serious c none Although the Bryden 1997 [22] study showed that Entonox (nitrous oxide plus oxygen) along with isoflurane inhalation was more effective for pain related to the first of 2 chest tubes to be removed, removal of the second tube was more painful, regardless of the gas inhaled. SMD -26.00 [-40.76, -11.24] on a 0 - 100 VAS.


CRITICAL

CI: Confidence interval; MD: Mean differenceExplanationsa. Unclear description of blinding practices in single included study. b. Comparator not placebo. c. Despite confidence intervals that don't cross no effect there was a small number of patients.

Question

Should Isoflurane & Nitrous Oxide vs. Nitrous oxide alone be used for critically ill adults undergoing a procedure (2.1)?


Background:

Intervention: Isoflurane & Nitrous Oxide

Comparison: Nitrous oxide alone


Assessment




Non-pharmacological interventions for procedural pain management are important given the documented prevalence of pain associated with many ICU procedures.



○ No included studies ● Very low




Pain Score


35 patients after uncomplicated cardiac surgery; chest tube removal. RCT double-blinded.Although the Bryden 1997 [22] study showed that Entonox (nitrous oxide plus oxygen) along with isoflurane inhalation was more effective for pain related to the first of 2 chest tubes to be removed, removal of the second tube was more painful, regardless of the gas inhaled. SMD -26.00 [-40.76, -11.24] on a 0 - 100 VAS.Are the desirable anticipated effects large?Although a SMD difference of 26 points on a 0 - 100 VAS is large, the results differed according to the order of the tube removal.Change to probably yes considering the change in pain scores and the difference between treatments.Gerald: Probably no because we should compare the first tube




○ No ● Probably no

removal - no difference between the 2 groups. Isoflurane given for the first tube, then it stopped. The patient awakened before the second tube removal. It's the effect of NOT having isoflurane for the second tube. The two tubes should be studied separately. The score difference is due to the study design and not real effects.Are the undesirable anticipated effects small? Patients inhaled gas until drowsy but not unconscious. Researchers noted no




○ No ● Probably no


● No Gases contained in high pressure cylindars. While they are a part of intra-operative anesthesia, use of these gases in ICU is unusual. Researchers noted that the the procedure was supervised by one of them (assume they are anesthesiologists?), the nurses administered the gas. Study done in Scotland, where nursing practice may differ from other places in the world.


● No Although the cost of this intervention was not addressed in the study, it is assumed that the cost would be large vis a vis the use of a drug analgesic. (Cost of equipment, cost of physician time.) More importantly, the effectiveness of this intervention for chest tube removal pain is only based on this one study.Very expensive; you need monitoring and safety procedures


● No For reasons noted above, in incremental costs.


● No For reasons noted above, in incremental costs.

RecommendationShould Isoflurane & Nitrous Oxide vs. Nitrous oxide alone be used for critically ill adults undergoing a procedure (2.1)?












○ ● ○ ○ ○





● ○ ○ ○

Recommendation We recommend not using inhalation of isoflurane for pain control during chest tube removal.Consider other less invasive, less costly, and more studied interventions for chest tube removal pain.

Justification The effectiveness of this intervention for chest tube removal pain is only based on this one study with a sample of 35 patients. Outcomes were not clearly reported.

Subgroup considerations The study was conducted in cardiac surgery patients.


See under, "Are the resources required small?", above.Not feasible to implement.

Monitoring and evaluation Evaluate patient pain before and after chest tube removal procedure as a factor of use of an analgesic intervention.

Research possibilities Pursue other types of analgesic interventions.

Question: A narcotic compared to an NSAID for critically ill adults undergoing a procedure (2.1) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other considerations a narcotic an

NSAIDRelative

(95% CI)

Absolute

(95% CI)

Pain Score

1 randomised trials

not serious

not serious not serious very serious a

none 17 19 - MD 0.45 lower(2.16

lower to 1.26

higher)

⨁⨁◯◯LOW

CRITICAL

CI: Confidence interval; MD: Mean differenceExplanations a. Wide confidence intervals don't exclude harm or benefit and small number of patients.

Question

Should a narcotic vs. an NSAID be used for critically ill adults undergoing a procedure (2.1)?


Background:


Comparison: an NSAID


Assessment



○ No ○ Probably no ○ Uncertain ○ Probably yes ● Yes



○ No included studies ○ Very low ● Low




Pain Score


One single study (Puntillo & Ley, 2004) [23]: 74 patients randomized in a double-blind study: 1) 4 mg of IV morphine + procedural information (n=17); 2) 30 mg of IV ketorolac + procedural information (n=19); 3) 4 mg of IV morphine + procedural + sensory information (n=19), and 4) 30 mg of IV ketorolac + procedural + sensory information (n=19). No control group included.Procedure: CRT in cardiac surgery ICU patients0-10 NRS pain intensity scores did not differ significantly across time among the 4 groups (RM-ANOVA F=0.82, p=0.49). Baseline pain scores (varied from 2.24 to 3.45), immediately after CTR (varied from 1.63 to 4.40), and 20 minutes after CTR (varied from 0.63-1.85). Self-reported pain scores appeared lower in Group 2 and Group 4 (but were not statistically significant). MD was 0.45 lower.Are the desirable effects large?Probably no because the MD was small. Effect size was also very small (0.03 for pain intensity).Are the undesirable anticipated effects small?Uncertain because they were not documented in this study.Are the desirable effects large relative to undesirable effects?Uncertainty because the desirable effects were small, and undesirable effects unknown.




● Probably no







Yes because IV morphine and IV ketorolac are part of standard of care, and easily accessible for administration.



Probably yes because both treatments are easily accessible and seem to have similar effects. However, pain intensity was mild in all groups.JF: Both treatments are inexpensive.Paul: Benefits are uncertain.Bram: Uncertain better in this case.



Yes because these drugs are already part of ICU standard of care.



Probably yes because to time the peak effect of medication with CTR can be difficult in practice (i.e., uncertainty about the time of CTR procedure).

RecommendationShould a narcotic vs. an NSAID be used for critically ill adults undergoing a procedure (2.1)?












○ ○ ● ○ ○





○ ○ ● ○

Recommendation We suggest offering either IV morphine or IV ketorolac prior to a procedure (i.e., CTR) for pain management in ICU patients (+2C).should we keep general terms (opioid, NSAID)?for procedures, can we extrapolate to any procedure?question of interest to clinicians (for any procedure).

Justification The study findings did not confirm the relative superiority of either morphine or ketorolac as an analgesic prior to CTR. Pain intensity was mild in all groups.

Subgroup considerations CTR in cardiac surgery ICU patients


Timing the administration of analgesia (peak effect) prior to the procedure is a key aspect to maximize effective pain reduction. Coordination with the care team is important (and can be challenging).


Pain assessment pre and post procedure to allow the evaluation of the effectiveness of analgesia.; monitoring of adverse events is also important to plan.

Research possibilities Include a control group and larger sample size; investigate other procedures and other medication (opioids, NSAID) in various ICU patient groups.Implementation studies are also needed


The recommendation’s wording differs from the question – realizing it’s hard to make firm statements with low quality of evidence. Please qualify for "patients with normal kidney function not at risk of AKI" Recommendation & e to d table inconsistent; e to d recommends narcotic not nsaid; Recommendation based on one study with ketorolac, generalizing it to all NSAIDS and all procedures.

Question: NSAID gel compared to placebo for critically ill adults undergoing a procedure (2.1) Setting: Bibliography:



s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision


NSAID gel placebo Relative

(95% CI)

Absolute

(95% CI)

Pain Score

1 randomised trials

not serious

not serious not serious serious a none 53 53 - MD 3 lower(3.56

lower to 2.44

lower)


CRITICAL

CI: Confidence interval; MD: Mean differenceExplanations a. Lowered for small number of patients despite tight confidence intervals showing benefit.

Question

Should NSAID gel vs. placebo be used for critically ill adults undergoing a procedure (2.1)?


Background:

Intervention: NSAID gel

Comparison: placebo


Assessment


Problem Is there a problem priority?● Yes


What is the overall certainty of this evidence? ● Moderate The relative importance or

values of the main outcomes of interest:

Outcome

Relative importa

nce

Certainty of the

evidence (GRADE)

Pain Score

CRITICAL


Singh study [24] : Doubleblind RCT of cardiac patients with mediastinal and pleural chest tubes. Control: paraffin gel around chest tubes (either mediastinal or pleural) (n=53) vs. 50 mg valdecoxib around chest tubes (either mediastinal or pleural) (n=53) at least 30 minutes prior to CTR (median 45 minutes).0-10 VAS scores: control 5 +1.48 vs valdecoxib 2 + 1.48 (p<0.05)SMD: -3 [-3.56, -2.44]

Desirable effects large? clinically significant differences in pain scoresUndesirable effects small? no differences in HR, BP and other "adverse effects" such as skin irritation.




● Yes


● Yes


● Yes

Resource use Are the resources required small? ● Varies

Application of NSAID around tube 30 minutes prior to CTR should not be too time intensive. Probably more costly than opioids, although authors said cost for 10 g Valdecoxib would cost US $0.50.- No gel available at the bedside in the US. - very expensive ($200/patient).-: How to address availability issues? To be discussed in a larger forum. US cost to consider?What is the cost of voltaren local analgesic gel?-: Varies based on availability issues in different countries.


● VariesNet benefits were large, availability and costs will vary.

Acceptability Is the option acceptable to key stakeholders? ● Varies

Stakeholders could accept depending on availability and cost.


○ No ○ Probably no ○ Uncertain ○ Probably yes ○ Yes ● Varies

With time, practice of applying NSAID gel would become routinized.-: Varies based on options available in hospitals.-: Initially Probably yes if available and at low cost, but varies is acceptable.

RecommendationShould NSAID gel vs. placebo be used for critically ill adults undergoing a procedure (2.1)?












○ ○ ○ ○ ●





○ ○ ● ○

Recommendation Study results show clinically significant effects on CTR pain. Application to practice is recommended.We suggest offering using NSAID gel (if available) over placebo prior to CTR.Small sample size + uncertainty, change for "we suggest offering this option". Vote to comeVery expensive product; no direct comparison besides placebo.NSAID gel available available over the counter in France (not in the USA - Paul).

Justification CTR pain is substantial and needs effective interventionOne study from India with a product not available in the US.

Subgroup considerations Study conducted in India; needs replication in heterogeneous group of patients elsewhere and in patients with chest tubes after thoracic surgery or trauma, and in larger sample sizes.

Implementation considerations Gel needs to be applied long enough prior to procedure (30 to 60 minutes) to get peak effect.Availability and costs to be taken into consideration.

Monitoring and evaluation Systematic assessment of pain scores before and during/after CTR.

Research possibilities See subgroup considerations. Other studies testing other products available in other countries.


The recommendation is inconsistent with the e to d profile, which recommends nsaid gel

Question: Cold therapy compared to no cold therapy for critically ill adults undergoing a procedure (2.2) Quality assessment № of patients Effect

Quality Importance

№ of studie

s

Study design

Risk of bias

Inconsistency

Indirectness

Imprecision

Other consideration

s

cold therap

y

no cold therapy

Relative(95% CI)

Absolute

(95% CI)

Pain Intensity during Procedure (assessed with: 2 mins post procedure)

2 randomised trials

not serious

not serious serious a serious b none 65 65 - MD 1.91 lower(5.34

lower to 1.52

higher)

⨁⨁◯◯LOW

CRITICAL

CI: Confidence interval; MD: Mean differenceExplanationsa. Multiple time points after chest tube removal studied and only 2 mins post was significantly different (these are results presented here). b. Small number of patients and confidence intervals do not rule out harm. Question

Should cold therapy vs. no cold therapy be used for critically ill adults undergoing a procedure (2.2)?


Background:

Intervention: cold therapy

Comparison: no cold therapy

Main outcomes: Pain Intensity during Procedure

Assessment




10/12/16Dear MA Heidari Gorji, I am a member of a Guidelines panel on Pain, Agitation , Delirium, Early Mobilization and Sleep for the Society of Critical Care Medicine. I read your report on ice packs for chest tube removal pain with great interest, and we would like to include it in our evidence results. However, we have some questions about your study that I hope you can answer. 1) Can you present the numbers in your sample? We are confused as to whether 80 post-cardiac surgery patients were in the study (abstract). Under Materials and Methods, you write that the sample size of 30 patients in each group (intervention and control) was increased to 40 in each group. In this same section, in the paragraph that begins with “On the first day after operation”, you describe assignment of patients to groups which is confusing to me. Under Statistical Analysis, you say there were 90 patients for whom data were available. Earlier you had said there were 80 patients. In Table 1, I count 40 (male and female) patients in the experimental group and 40 (male and female)patients in the control group. Regarding Table 2, how many patients were in each of the three groups: control, cold application, and relaxation?

2) Regarding the pain measurement tool that you used: Under Materials and Methods, in the paragraph that begins with “Data collection tools…”, you write that you used a 0 – 100 mm Visual Analogue Scale. However, in Table 2, you report pain severity scores as 4.6+0.66 (control) vs 2.4+0.52 (cold application) vs 2.3+0.35 (relaxation). Are those scores on a 0 – 100 VAS or a 0 – 10 numeric rating scale (NRS)? In Figure 1, your “y” axis is 0 – 5. Was that per a 0 – 100 VAS or a 0 – 10 NRS?3) Do you provide differences in pain scores according to the type of chest tube removed, mediastinal or pleural?

Thank you for your help with this.




○ No included studies ○ Very low ● Low ○ Moderate ○ High




Pain Intensity during Procedure


Overall Certainty: 2 studies were in data abstraction. Gorji 2014 [25] was true RCT and Sauls 2002 [26] was before and after study. The results were different (Gorji showed benefit whereas Sauls did not). In Gorji[25] , for intervention group #1, iced packs wound around chest tube x 10 minutes until temperature 13 degrees; for intervention group #2, relaxation: calm and deep breathing for 5 minutes, then 10 minutes more; control group#3 received no intervention. Used 0 - 100 VAS but must have reported on a 0 - 10 scale.Pain intensity scores immediately after CTR:cold application group = 2.4 + 0.52control group = 4.6 + 0.66Size of each of the groups unknown. See email.Gorji: SMD 3.67 [-4.40, -2.94] lower for cold therapy; SignificantIn Sauls[26] , for intervention group, ice pack on either side of chest tube for 10 minutes; for control group, tepid water pack on either side of chest tube. Used 0 - 10 NRS.Pain intensity scores immediately after CTR:Control group (n=25): 6.34 + 2.52experimental group (n=25): 5.86 + 2.82Sauls: SMD -0.17 lower for cold therapy [-0.72, 0.39] NSAre the desirable anticipated effects large?The overall SMD is -1.91 [-5.34, 1.52]Therefore, the desirable anticipated effects are


○ Important uncertainty or variability ○ Possibly important uncertainty or variability ○ Probably no important uncertainty or variability ● No important uncertainty or variability





probably large because a NRS pain intensity difference greater than 1.7 - 2 is assessed as clinically significant. However, this was from only 1 study with a questionable sample size.Group consensuAre the undesirable anticipated effects small? There was no mention in either study of side effects associated with the use of cold packs. But it is possible that patients could have skin burns





No description of resources besides the gel packs in studies. Three cooling gel packs (8 × 10 cm) twisted in gauze were used to wrap around each chest tube for 10 minutes prior to tube removal. These packs would have to be purchased and stored in a freezer in the ICU. Having enough room for more than one set may pose a problem.



The overall costs of the gel cooling packs should be incidental if the benefits to patients were decreased pain associated with a procedure known to cause substantial pain. Patients who had received opioid analgesic during less than 4 hours before the intervention were excluded from the study. However, costs of an analgesic intervention, if needed to augment cold therapy, would be low.



Key stakeholders should be committed to a relatively low cost, non-pharmacological approach to pain relief.



Implementation would require a practice change and development of a written protocol for use of the gel packs, as well as purchase and storing of the gel packs.-Uncertain: only 2 studies with contradictory results. In one study, patients did not receive opioids.- Probably yes; question on feasibility of implementing of the intervention.- Feasible to implement considering the low amount of resources needed compared to other interventions.

RecommendationShould cold therapy vs. no cold therapy be used for critically ill adults undergoing a procedure (2.2)?












○ ○ ○ ● ○





○ ○ ● ○

Recommendation We suggest to offer cold therapy during chest tube removal for ICU pain management (to be confirmed).More research is needed, with inclusion of larger samples, before the use of cold gel packs is determined to be efficacious for pain relief during chest tube removal.Note: Still waiting to hear from Gorji to answer our questions.

Justification Chest tube removal pain was decreased in one study of cold packs.

Subgroup considerations These 2 studies were conducted with cardiac surgery patients. In Gorji study, patients had 1 mediastinal tube and 1 left pleural tube. Differences in pain between the 2 types of tubes were not reported. In previous research, (Puntillo & Weiss, 1994) [27], pleural chest tubes were significantly more painful than mediastinal tubes. So, type of tube should be analyzed.


Implementation would require a practice change in a particular ICU.


Pre and post pain assessments to evaluate impact of the intervention. Patient's preference and satisfaction to be considered.

Research possibilities The Gorji [25] study was conducted in Iran. Other patient groups, such as those with chest tubes after thoracic surgery or after trauma, should be studied. Mechanically ventilated patients were excluded from the Gorji [25] study. Those patients could be included in future studies and, if they are unable to self-report pain, a pain behavior scale could be used to assess pain. These steps could increase generalizability of study results. If patients reported pain prior to the procedure, they could be medicated with analgesic drugs, and the effect of these drugs, as well as the ice, could be examined


Unclear whether cold/ music therapy offered as sole pain management strategy or considered adjuncts from the question.Evidence poor; RCT study= beneficial, pre-post negative, gaps in methods/ results and methods.

Question: Relaxation with pharmacotherapy compared to no relaxation for critically ill pain management (5.4) Quality assessment № of patients Effect

Quality

Importance

№ of studie

s

Study design

Risk of

bias

Inconsistency

Indirectness

Imprecision

Other consideratio

ns

Relaxation with pharmacothera

py

no relaxatio

n

Relative

(95% CI)

Absolute

(95% CI)

Pain Intensity immediately after Chest Tube Removal (VAS cm)


not serious a

serious b,c not serious very serious d,e

none 31 33 - MD 0.32

higher(0.67

lower to 1.31

higher)

⨁◯◯◯

VERY LOW

CRITICAL

Pain intensity 15 minutes after Chest tube removal (VAS cm)


not serious

not serious not serious serious d,f none 19 21 - MD 2.5 lower(4.18

lower to 0.82

lower)

⨁◯◯◯

VERY LOW

CRITICAL

CI: Confidence interval; MD: Mean differenceExplanationsa. 1 study indicated low risk of bias; 2nd study indicated high ROB due to variable co-interventions between groups and no adjustment for group differences b. 1 study showed variable results based on gender and age; results not consistent amongst groups c. I2 = 80% d. Low sample size e. mean difference crosses line of no effect f. Despite CI that do not cross 'no effect', imprecision lowered for sample size

Question

Should Relaxation with pharmacotherapy vs. no relaxation be used for critically ill pain management (5.4)?

Population: critically ill pain management (5.4)

Background:

Intervention: Relaxation with pharmacotherapy

Comparison: no relaxation

Main outcomes: Pain Intensity immediately after Chest Tube Removal (VAS cm)


Setting:

Perspective:

Assessment







○ No included studies ● Very low

Summary of overall evidence on pain scores:2 matched cohort studies (Friesner et al. 2006; Houston & Jesurum, 1999) [28, 29] in which pain scores were higher


OutcomeRelative

importance

Certainty of the

evidence (GRADE

)

Pain Intensity immediately after Chest Tube Removal (VAS cm)




immediately after CTR with a MD of 0.32, and then decreased with a MD of 2.5 15 minutes after CTR. However, the quality of evidence was very low. Small samples were used in both studies (n=40 and 24, respectively).In Friesner et al. (2006) (n=40), relaxation consisted of deep-breathing exercises. Immediately after CTR, mean pain scores were 6.57 (SD=2.61) for the intervention group, and 8.61 for the control group (SD=2.96). Mean pain scores decreased to 3.07 (SD=2.45) and to 5.57 (SD=2.96) in the intervention group, and the control group respectively. Significant decreases in pain scores with relaxation were found immediately after, and 15 minutes post CTR (F tests, p<0.01) (p. 273).In Houston & Jesurum (1999) [29] (n=24), the quick relaxation technique (QRT) consisted of clenching fists, deep breathing, go limp as a rag doll, and yawning. Strange results were obtained and were presented by age and gender. In summary, pain scores were not significantly different between groups (p=0.062). Men older than 70 yo in the control group showed almost twice more pain than those of the intervention group. The situation was reversed in older women (p.202).Are the desirable anticipated effects large?Likely uncertain, 2 non RCTs with small samples and very low quality of evidence. Opioid administration was variable between patients.Are the undesirable anticipated effects large?No mention of adverse events in both studies.Are the desirable effects large relative to undesirable effects?Uncertain based on very low evidence and no information on adverse events (although they appear to be very unlikely).






● Probably yes





Instructions clearly described in both studies.

Short instructions which can be taught to ICU patients at the bedside.



Is relaxation worth the benefit?The net benefits remain to be demonstrated, but appear promising. Relaxation does not involve any cost.



I would believe that the vast majority would likely approve this intervention.Patient's preference to take into consideration.



No implementation studies available.In Friesner et al. (2006)[28] , the relaxation techinique was initiated by patients 5 minutes before CTR.In Houston & Jesurum (1999)[29] , the quick relaxation technique was taught to patients the day prior to CTR, and participated in two practice sessions (one immediately prior to CTR).

Little resource and time needed to perform a brief relaxation technique during procedural pain.Minimal training required.Nursing training? Will depend on relaxation technique.

RecommendationShould Relaxation with pharmacotherapy vs. no relaxation be used for critically ill pain management (5.4)?












○ ○ ● ○ ○


We recommend against offering this option




○ ○ ● ○

Recommendation We suggest offering relaxation during procedural pain as a complementary intervention to ICU pain management (+2C).Can we still suggest an intervention if the benefits and consequences are balanced or uncertain? Do we change to probably outweigh?

Justification Although evidence is of very low quality, relaxation remains a safe and quick intervention to use. However, the benefits remain to be demonstrated. Worth mentioning, various relaxation techniques were used in included studies.

Subgroup considerations Both included studies were performed in cardiac surgery (CABG) ICU patients during CRT (chest tube removal).


No specific resources needed to apply this intervention. Minimal training required. Written information could also be provided to patients to get familiar with relaxation.


Pre and post pain assessments to evaluate impact of the intervention. Patient's preference and satisfaction could also be considered.

Research possibilities Experimental studies (RCT) are clearly needed to demonstrate the benefits of relaxation. Implementation studies are also lacking.

Question: Standardized and/or protocol-based analgesia/analgosedation programs compared to no standardized programs for critical care pain management (4) Setting: Critical Care ICU Bibliography:



s

Study design

Risk of

bias

Incon-sistency

Indirect-ness

Imprec-ision


standardized and/or protocol-based

analgesia/analgosedation programs

no standardized programs

Relative

(95% CI)

Absolute

(95% CI)

Nosocomial infection

2 randomised trials

serious a

not serious

not serious b

serious c,d none 6/112 (5.4%) 10/106 (9.4%)

RR 0.62

(0.25 to 1.56)

36 fewer per

1,000(from

53 more to 71

fewer)

⨁⨁◯◯LOW

CRITICAL

Duration of MV

3 randomised trials

serious e

not serious

not serious

not serious

none 313 326 - MD 1.26

lower(1.8

lower to 0.73

lower)

⨁⨁⨁◯MODERAT

E

CRITICAL

ICU LOS

3 randomised trials

serious e

not serious

not serious

not serious

none 313 326 - MD 2.27

lower(2.96

lower to 1.58

lower)

⨁⨁⨁◯MODERAT

E

CRITICAL

Constipation


not serious

not serious

not serious

serious d none 18/50 (36.0%) 15/50 (30.0%) RR 1.20(0.68 to

2.11)

60 more per 1,000

(from 96 fewer to 333

more)


IMPORTANT



s

Study design

Risk of

bias

Incon-sistency

Indirect-ness

Imprec-ision


standardized and/or protocol-based

analgesia/analgosedation programs

no standardized programs

Relative

(95% CI)

Absolute

(95% CI)

Cardiovascular Adverse Drug Reaction

2 randomised trials

serious f

not serious

not serious b

serious d none 4/153 (2.6%) 4/157 (2.5%) RR 0.90

(0.26 to 3.14)

3 fewer per 1,000(from 19 fewer to

55 more)

⨁⨁◯◯LOW

CRITICAL

Sedatives Exposure

1 randomised trials

serious g

not serious

not serious

serious c none 110 116 - SMD 0.57 lower(0.84 lower to 0.31

lower)

⨁⨁◯◯LOW

CRITICAL

Opioid Exposure

1 randomised trials

serious g

not serious

not serious

serious c none 55 58 - MD 0.06 SD higher


⨁⨁◯◯LOW

CRITICAL

Pain intensity VAS (cm)


not serious

not serious

not serious

not serious

none 747 780 - MD 0.35 lower(0.49 lower to 0.22

lower)

⨁⨁◯◯LOW

CRITICAL

CI: Confidence interval; RR: Risk ratio; MD: Mean difference; SMD: Standardised mean differenceExplanationsa. 1 study reports no blinding of participants or outcome assessors; the other study does not report randomization or blinding processes b. Note: only includes remifentanil based analgosedation protocols c. Small sample size d. Unclear if benefit or harm e. Two studies report randomization; 1 study did not report randomization; no studies had adequate blinding of participants or outcome assessors f. 1 study did not clearly report randomization or blinding of participants/assessors; 2nd study did not randomize or blind participants/assessors g. Appropriate randomization and intention-to-treat, however inadequate or lack of randomization

Question

Should standardized and/or protocol-based analgesia/analgosedation programs vs. no standardized programs be used for critical care pain management (4)?

Population: critical care pain management (4) Background:

Intervention: standardized and/or protocol-based analgesia/analgosedation programs

Comparison: no standardized programs

Main outcomes: Nosocomial infection Duration of MV ICU LOS Constipation Cardiovascular Adverse Drug

Reaction Sedatives Exposure Opioid Exposure Pain intensity VAS (cm)

Setting: Critical Care ICU

Assessment



● Yes



○ No included studies ○ Very low ● Low


OutcomeRelative importan

ce

Certainty of the evidence (GRAD

E)

are the desirable anticipated effects large?reduction in duration of MV and reduction in ICU LOS(data from 3 RCTs)Duration of MV = MD - 1.26 (-1.8 to -0.73)ICU LOS = MD -2.27 (-2.96 to -1.58)Rozendaal FW, Spronk PE, Snellen FF, et al: Remifentanil-propofol analgo-sedation shortens duration of ventilation and



length of ICU stay compared to a conventional regimen: A centre randomised, cross-over, open-label study in the netherlands. Intensive Care Med 2009;35:291-298 [30]Brook AD, Ahrens TS, Schaiff R, et al: Effect of a nursing-implemented sedation protocol on the duration of mechanical ventilation. Crit Care Med 1999;27:2609-2615 . [31]Strom T, Martinussen T, Toft P: A protocol of no sedation for critically ill patients receiving mechanical ventilation: A randomised trial. Lancet 2010;375:475-480

Pain intensity - "reduced" 4 cohort trials NRS (0-10) - pooarburns. Burns 2010;36:639-646 [32]undergoing cardiac surgery after implementation of a quality improvement postoperative pain treatment program. J Crit Care 2008;23:359-371 [33] Egerod I, Jensen MB, Herling SF, et al: Effect of an analgo-sedation protocol for neurointensive patients: A two-phase interventional non-randomized pilot study. Crit Care 2010;14:R71 [34]Skrobik Y, Ahern S, Leblanc M, et al: Protocolized intensive care unit management of analgesia, sedation, and delirium improves analgesia and subsyndromal delirium rates. Anesth Analg 2010;111:451-463 [35]nosocomial infection - no difference combining 2 RCTRR 0.62 (0.25 to 1.56)36 fewer per 1000 patientsBreen D, Karabinis A, Malbrain M, et al: Decreased duration of mechanical ventilation when comparing analgesia-based sedation using remifentanil with standard hypnotic-based sedation for up to 10 days in intensive care unit patients: A randomised trial [ISRCTN47583497]. Crit Care 2005;9:R200-10 [36] Strom T, Martinussen T, Toft P: A protocol of no sedation for critically ill patients receiving mechanical ventilation: A randomised trial. Lancet 2010;375:475-480) [37]Discussion:JF: Yes considering effects on MV duration and ICU LOSCeline + Kathleen: Probably yes considering no clinical significance on pain scores (refer to question)Gerald: Yes - protocol with sedatives which impact on patient's LOC so more impact on other outcomes than pain intensityUse of sedatives and analgesics will influence pain scores - Berger trial (1.4 control vs 0.9 intervention), 2.7 control vs 1.4 protocol, 1.61 control vs 1.15 protocol, 1.5 vs 0.6So pain scores were low (average over the ICU LOS) in both groups and the differences was still significant - could change to


● Yes


● Probably yes ○ Yes ○ Varies



Yes based on this informationare the undesirable anticipated effects small?constipation - no difference(although only 1 cohort trial assessed it - Bowel movement at 48 hours)RR 1.20 (0.68 to 2.11)60 more per 1000 ptTedders KM, McNorton KN, Edwin SB: Efficacy and safety of analgosedation with fentanyl compared with traditional sedation with propofol. Pharmacotherapy 2014;34:643-647Cardiovascular ADR no difference (2RCT)RR 0.90 (0.26 to 3.14)3 fewer per 1000 ptBreen D, Karabinis A, Malbrain M, et al: Decreased duration of mechanical ventilation when comparing analgesia-based sedation using remifentanil with standard hypnotic-based sedation for up to 10 days in intensive care unit patients: A randomised trial



can be a big practice change to end users. May require more resourcesimplementing protocols can be time consuming and maintaining the quality is challenging: agree: Probably no or varies - resources are variable from an ICU to another



Reduction in ICU LOS drives economics, however it is important to note the implementation typically requires dedicated support.there MAY be an increase need for 1:1 nursing care (or use of sitters - leave out?)In addition, it is important to note some of the cohort trials and some of the RCT utilize remifentanil. This question does not cull out individual medications; therefore the cost of specific medication should not influence cost.comment on use of sitters (could be seen as a cost barrier) - may be related to more agitation less sedatives, more awake patients may lead to increased presence at the bedsideIn some countries the ratio of nurse-pt is 1:1 (nurse also manage the ventilator); more mobilization too specific to one study



Standardized and/or protocol-based analgesia/analgosedation programs can be polarizing topic. Some multi-professional ICU leaders are very supportive and others are not.Change to Probably yes - although we know that not all opioids are easily available in all countries-Nurse satisfaction increased after implementation of such a protocol – prob.yes

Feasibility Is the option feasible to implement? ● Probably yes

it's been done in several real-life cohort examples, therefore likely feasible (but not easy) Group agrees!

RecommendationShould standardized and/or protocol-based analgesia/analgosedation programs vs. no standardized programs be used for critical care pain management (4)?












○ ○ ○ ● ○





○ ○ ● ○

Recommendation We suggest standardized and/or protocol-based analgesia/analgosedation programs should be utilized for critical care pain management. (+2C)*3 RCTs with consistent results; suggest to change with "recommend" (Kathleen agrees)*refer to 2013 guidelines (suggest) because it was a +2B*C for low quality of evidence so should be +2C*Recommend has a stronger impact on policy makers - it would require that all ICUs have protocols for their patients - does not mean that pain not be managed it can be done individually*agrees with "recommend" but also with consistency with previous guidelinesPaul: stay with "suggest"*analgosedation defined as "analgosedation refers to the use of opioids for pain control and sedation purposes." by the group

Justification Reduction in ICU LOS and MV with no apparent increase in ADE (adverse drug events) compared to standard of care.

Subgroup considerations Many subgroups have been studied, however much more research in need (see research possibilities below) in specific ICU population. Neuro ICU, Trauma ICU, Cardiac Surgery ICU.


Institutions should have have frequent assessment (multiple times a day) of pain via validated tools as a standard of care (patient report and behavioral pain scores). Leaders need to be willing to champion the effort to embrace the challenges of implementing clinical practice change.


Important to monitor process measures (e.g. quality of pain assessment documentation, compliance with frequency, opioid/sedation mediation utilization) and clinical outcomes (e.g. pain scores, ICU LOS, length of MV) as well as economic impact.

Research possibilities is one opioid better than another for standardized and/or protocol-based analgesia/analgosedation programs ? Gerald - clearly define analgesia-based sedation Is there an ICU setting (ICU patient population) most appropriate for standardized and/or protocol-based analgesia/analgosedation programs?Is the benefit of standardized and/or protocol-based analgesia/analgosedation programs reduction in pain or is it avoidance of harmful sedatives?More in depth look at the adverse effects, opioid withdrawal and post-hospital opioid use disorder associated with standardized and/or protocol-based analgesia/analgosedation programs.


Hesitant about strong recommendation; moderate evidence on ICU LOS and MV duration. With validated pain assessment tools available, it is time to move forward and implement structured pain management approaches. Comparative group needs to be added into the recommendation. "Assessment driven" is not in the question; qualifier warrants justification

References

1. Kim K, Kim W-J, Choi D-K, Lee YK, Choi I-C, Sim J-Y: The analgesic efficacy and safety of nefopam in patient-controlled analgesia after cardiac surgery: A randomized, double-blind, prospective study. J Int Med Res 2014, 42(3):684-692.

2. Cattabriga I, Pacini D, Lamazza G, Talarico F, Di Bartolomeo R, Grillone G, Bacchi-Reggiani L: Intravenous paracetamol as adjunctive treatment for postoperative pain after cardiac surgery: a double blind randomized controlled trial. Eur J Cardiothorac Surg 2007, 32(3):527-531.

3. Memis D, Inal MT, Kavalci G, Sezer A, Sut N: Intravenous paracetamol reduced the use of opioids, extubation time, and opioid-related adverse effects after major surgery in intensive care unit. J Crit Care 2010, 25(3):458-462.

4. Guillou N, Tanguy M, Seguin P, Branger B, Campion J-P, Mallédant Y: The effects of small-dose ketamine on morphine consumption in surgical intensive care unit patients after major abdominal surgery. Anesth Analg 2003, 97(3):843-847.

5. Wang L, Johnston B, Kaushal A, Cheng D, Zhu F, Martin J: Ketamine added to morphine or hydromorphone patient-controlled analgesia for acute postoperative pain in adults: a systematic review and meta-analysis of randomized trials. Can J Anaesth 2016, 63(3):311-325.

6. Assouline B, Tramèr MR, Kreienbühl L, Elia N: Benefit and harm of adding ketamine to an opioid in a patient-controlled analgesia device for the control of postoperative pain: systematic review and meta-analyses of randomized controlled trials with trial sequential analyses. Pain 2016, 157(12):2854-2864.

7. Pandey CK, Bose N, Garg G, Singh N, Baronia A, Agarwal A, Singh PK, Singh U: Gabapentin for the treatment of pain in guillain-barré syndrome: a double-blinded, placebo-controlled, crossover study. Anesth Analg 2002, 95(6):1719-1723, table of contents.

8. Pandey CK, Raza M, Tripathi M, Navkar DV, Kumar A, Singh UK: The comparative evaluation of gabapentin and carbamazepine for pain management in Guillain-Barré syndrome patients in the intensive care unit. Anesth Analg 2005, 101(1):220-225, table of contents.

9. Pesonen A, Suojaranta-Ylinen R, Hammarén E, Kontinen VK, Raivio P, Tarkkila P, Rosenberg PH: Pregabalin has an opioid-sparing effect in elderly patients after cardiac surgery: a randomized placebo-controlled trial. Br J Anaesth 2011, 106(6):873-881.

10. Joshi SS, Jagadeesh AM: Efficacy of perioperative pregabalin in acute and chronic post-operative pain after off-pump coronary artery bypass surgery: a randomized, double-blind placebo controlled trial. Ann Card Anaesth 2013, 16(3):180-185.

11. Liu J, Wang L-N, McNicol ED: Pharmacological treatment for pain in Guillain-Barré syndrome. Cochrane Database Syst Rev 2013, 10(10).12. Eipe N, Penning J, Yazdi F, Mallick R, Turner L, Ahmadzai N, Ansari MT: Perioperative use of pregabalin for acute pain—a systematic review and meta-

analysis. Pain 2015, 156(7):1284-1300.13. Insler SR, O'Connor M, Samonte AF, Bazaral MG: Lidocaine and the inhibition of postoperative pain in coronary artery bypass patients. J Cardiothorac

Vasc Anesth 1995, 9(5):541-546.14. Kranke P, Jokinen J, Pace NL, Schnabel A, Hollmann MW, Hahnenkamp K, Eberhart LHJ, Poepping DM, Weibel S: Continuous intravenous

perioperative lidocaine infusion for postoperative pain and recovery. Cochrane Database Syst Rev 2015(7):CD009642.15. Oberhofer D, Skok J, Nesek-Adam V: Intravenous ketoprofen in postoperative pain treatment after major abdominal surgery. World J Surg 2005,

29(4):446-449.16. Mosso-Vázquez JL, Gao K, Wiederhold BK, Wiederhold MD: Virtual reality for pain management in cardiac surgery. Cyberpsychol Behav Soc Netw

2014, 17(6):371-378.17. Robleda G, Roche-Campo F, Sendra MÀ, Navarro M, Castillo A, Rodríguez-Arias A, Juanes-Borrego E, Gich I, Urrutia G, Nicolás-Arfelis JM et al:

Fentanyl as pre-emptive treatment of pain associated with turning mechanically ventilated patients: a randomized controlled feasibility study. Intensive Care Med 2016, 42(2):183-191.

18. Casey E, Lane A, Kuriakose D, McGeary S, Hayes N, Phelan D, Buggy D: Bolus remifentanil for chest drain removal in ICU: a randomized double-blind comparison of three modes of analgesia in post-cardiac surgical patients. Intensive Care Med 2010, 36(8):1380-1385.

19. Ahlers SJGM, van Gulik L, van Dongen EPA, Bruins P, van de Garde EMW, van Boven WJ, Tibboel D, Knibbe CAJ: Efficacy of an intravenous bolus of morphine 2.5 versus morphine 7.5 mg for procedural pain relief in postoperative cardiothoracic patients in the intensive care unit: a randomised double-blind controlled trial. Anaesth Intensive Care 2012, 40(3):417-426.

20. Barr J, Fraser GL, Puntillo K, Ely EW, Gelinas C, Dasta JF, Davidson JE, Devlin JW, Kress JP, Joffe AM et al: Clinical practice guidelines for the management of pain, agitation, and delirium in adult patients in the Intensive Care Unit: executive summary. Am J Health Syst Pharm 2013, 70(1):53-58.

21. Akrofi M, Miller S, Colfar S, Corry PR, Fabri BM, Pullan MD, Russell GN, Fox MA: A randomized comparison of three methods of analgesia for chest drain removal in postcardiac surgical patients. Anesth Analg 2005, 100(1):205-209.

22. Bryden FM, McFarlane H, Tunstall ME, Ross JA: Isoflurane for removal of chest drains after cardiac surgery. Anaesthesia 1997, 52(2):173-175.23. Puntillo KA, Morris AB, Thompson CL, Stanik-Hutt J, White CA, Wild LR: Pain behaviors observed during six common procedures: results from

Thunder Project II. Crit Care Med 2004, 32(2):421-427.24. Singh M, Gopinath R: Topical analgesia for chest tube removal in cardiac patients. J Cardiothorac Vasc Anesth 2005, 19(6):719-722.25. Gorji HM, Nesami BM, Ayyasi M, Ghafari R, Yazdani J: Comparison of Ice Packs Application and Relaxation Therapy in Pain Reduction during Chest

Tube Removal Following Cardiac Surgery. N Am J Med Sci 2014, 6(1):19-24.26. Sauls J: The use of ice for pain associated with chest tube removal. Pain Manag Nurs 2002, 3(2):44-52.27. Puntillo K, Weiss SJ: Pain: its mediators and associated morbidity in critically ill cardiovascular surgical patients. Nurs Res 1994, 43(1):31-36.28. Friesner SA, Curry DM, Moddeman GR: Comparison of two pain-management strategies during chest tube removal: relaxation exercise with opioids and

opioids alone. Heart Lung 2006, 35(4):269-276.29. Houston S, Jesurum J: The quick relaxation technique: effect on pain associated with chest tube removal. Appl Nurs Res 1999, 12(4):196-205.30. Rozendaal FW, Spronk PE, Snellen FF, Schoen A, van Zanten ARH, Foudraine NA, Mulder PGH, Bakker J, Ulti Si: Remifentanil-propofol analgo-

sedation shortens duration of ventilation and length of ICU stay compared to a conventional regimen: a centre randomised, cross-over, open-label study in the Netherlands. Intensive Care Med 2009, 35(2):291-298.

31. Brook AD, Ahrens TS, Schaiff R, Prentice D, Sherman G, Shannon W, Kollef MH: Effect of a nursing-implemented sedation protocol on the duration of mechanical ventilation. Crit Care Med 1999, 27(12):2609-2615.

32. Berger MM, Davadant M, Marin C, Wasserfallen J-B, Pinget C, Maravic P, Koch N, Raffoul W, Chiolero RL: Impact of a pain protocol including hypnosis in major burns. Burns 2010, 36(5):639-646.

33. Diby M, Romand J-A, Frick S, Heidegger CP, Walder B: Reducing pain in patients undergoing cardiac surgery after implementation of a quality improvement postoperative pain treatment program. J Crit Care 2008, 23(3):359-371.

34. Egerod I, Jensen MB, Herling SF, Welling K-L: Effect of an analgo-sedation protocol for neurointensive patients: a two-phase interventional non-randomized pilot study. Crit Care 2010, 14(2):R71.

35. Skrobik Y, Ahern S, Leblanc M, Marquis F, Awissi DK, Kavanagh BP: Protocolized intensive care unit management of analgesia, sedation, and delirium improves analgesia and subsyndromal delirium rates. Anesth Analg 2010, 111(2):451-463.

36. Breen D, Karabinis A, Malbrain M, Morais R, Albrecht S, Jarnvig I-L, Parkinson P, Kirkham AJT: Decreased duration of mechanical ventilation when comparing analgesia-based sedation using remifentanil with standard hypnotic-based sedation for up to 10 days in intensive care unit patients: a randomised trial [ISRCTN47583497]. Crit Care 2005, 9(3):R200-210.

37. Strøm T, Martinussen T, Toft P: A protocol of no sedation for critically ill patients receiving mechanical ventilation: a randomised trial. Lancet 2010, 375(9713):475-480.

Documents

Explanations - cdn-links.lww.com€¦ · Web viewNEW ZEALAND - TABLETS ONLY. OMAN. PAKISTAN. PANANA. QATAR. SALVADOR. SINGAPOUR - TABLETS ONLY. SWITZERLAND. THE BARBADOS - TABLETS